- Email: [email protected]
Quantitative assessment of neuromotor function in workers with current low exposure to mercury vapor
NeuroToxicology 29 (2008) 596–604
Contents lists available at ScienceDirect
NeuroToxicology
Quantitative assessment of neuromotor function in workers with current low exposure to mercury vapor Gunilla Wastensson a,*, Daniel Lamoureux b, Gerd Sa¨llsten a, Anne Beuter b,c, Lars Barrega˚rd a a
Department of Occupational and Environmental Medicine, Sahlgrenska University Hospital, and University of Go¨teborg, Box 414, SE-405 30 Go¨teborg, Sweden Centre de Neuroscience de la Cognition, Universite´ du Que´bec a` Montre´al, Montre´al, Que´bec, C.P. 8888, Canada H3C 3P8 c Institut de Cognitique, Universite´ Victor Segalen Bordeaux 2, Zone Nord, 146 rue Leo Saignat, 33076 Bordeaux Cedex, France b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 14 November 2007 Accepted 5 March 2008 Available online 20 March 2008
Evaluation of neuromotor function has been used in several epidemiological studies of workers with long-term exposure to mercury vapor (Hg0). Some recent studies indicate adverse effects at relatively low exposure levels. In the present study, we used sensitive quantitative methods, developed specifically to detect subtle effects of exposure to toxins on motor function. After exclusion of individuals with neurological diseases or other conditions that may affect performance, 43 chloralkali workers with current low exposure to Hg0, and 22 age-matched referents remained for further analysis. The median urinary mercury concentration in exposed workers was 5.9 mg/g (range 1.3–25) creatinine (mg/gC), while that in referents was 0.7 mg/gC (range 0.2–4.1). The mean exposure time was 15 years, and the median cumulative mercury index was 161 years mg/gC in exposed workers. A eurythmokinesimeter (EKM) was used to quantify eye–hand coordination, and a diadochokinesimeter, to measure rapid alternating rotation of the forearms. In general, the differences in performance between the exposed workers and the referents were small. Age was associated with a decrease in speed, more tremor, and longer contact duration between the stylus and the metal targets in performance of rapid pointing movements. Smokers had significantly more tremor, and more contacts per event in the EKM test, than nonsmokers. Taking age, shift work, and smoking habits into account, no significant associations with current or cumulative mercury exposure were found for the majority of the outcome variables from the quantitative tests. In general, this study indicates no significant adverse effects of Hg0 on neuromotor function at the exposure levels studied. ß 2008 Elsevier Inc. All rights reserved.
Keywords: Neuromotor function Mercury vapor Quantitative methods
1. Introduction Exposure to mercury vapor (Hg0) may cause adverse effects on many organs, including the central nervous system (CNS). The major routes of human exposure are inhalation of Hg0 from occupational sources, and dental amalgam (Clarkson and Magos, 2006). About 80% of the inhaled vapor is absorbed through the lungs. The highly lipophilic vapor spreads throughout the body dissolved in the blood, and readily crosses the blood–brain barrier (WHO, 2003). Inside the cells, Hg0 is oxidized to divalent mercury (Hg2+), which is assumed to be the proximate toxic agent (Clarkson and Magos, 2006). After exposure, most of the mercury in the brain is cleared with a short half-time, but a fraction may have a much longer half-time, of several years (WHO, 2003). Excretion is via urine and feces, with a whole-body, half-time of about 60 days.
* Corresponding author. Tel.: +46 31 786 2894; fax: +46 31 40 97 28. E-mail address: [email protected] (G. Wastensson). 0161-813X/$ – see front matter ß 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2008.03.005
Measurement of mercury in urine is considered to be the best indicator of body burden of mercury with long-term exposure, and is commonly used as biomarker of exposure (WHO, 2003). There is a wide range of cognitive, personality, sensory, and motor disturbances reported after exposure to Hg0 including tremor, emotional lability, polyneuropathy, and impaired performance in tests of cognitive function and psychomotor skills (WHO, 2003). Even if little is known about the exact pattern of distribution in the CNS among humans (Clarkson and Magos, 2006), the extent and variety in neuropsychological impairment following Hg0 exposure imply that most structures in the CNS are affected. One of the earliest sign of mercury intoxication is a typical intentional tremor, indicating impairment of the cerebellum, a structure involved in coordination and voluntary movements. Traditionally, the finger–nose test was used in periodic examinations of mercury-exposed workers; however, decreasing exposure levels due to improved hygienic conditions require more precise and sensitive tools. Some studies using sensitive tests show a possible effect on psychomotor skills, such as reduced motor speed
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
(Ngim et al., 1992; Echeverria et al., 1998; Lucchini et al., 2002), and tremor (Chapman et al., 1990; Fawer et al., 1983; Langworth et al., 1992) even at relatively low exposure levels. Moreover, meta-analyses performed on studies of mercury-exposed workers have shown a larger impairment on motor performance compared with other neuropsychological domains such as memory and attention (Meyer-Baron et al., 2004; Rohling and Demakis, 2006). The aim of this study was to investigate effects on neuromotor function in chloralkali workers with current low level exposure to Hg0 using new sensitive quantitative methods. The eurythmokinesimeter (EKM) used in the present study quantifies rapid and precise proximo-distal movements in a pointing task supposed to be similar to the finger–nose test. The ability to perform rapid alternating movements (pronation/supination of the forearms), diadochokinesis, is quantitatively measured by the diadochokinesimeter (DIADO). The reliability and the sensitivity of these techniques have been evaluated in populations exposed to methylmercury (Beuter et al., 1999a, 1999b) and manganese (Beuter et al., 1994; Beuter et al., 1999), and among subjects with Parkinson’s disease and other neurological deficits (Beuter et al., 1994, 1999a, 1999b; Fimbel et al., 2005).
2. Methods 2.1. Subjects We recruited 58 mercury-exposed workers and 35 unexposed referents, all men, from two different chloralkali plants located in the same geographic area of the southwest region of Sweden. The selection procedure has been described elsewhere in detail (Wastensson et al., 2006). Information concerning relevant background variables, and the subjects’ state of health was collected in a questionnaire and by an additional interview before a physical examination was performed. Individuals with diseases, medication, or other circumstances that could possibly affect neuromotor function were excluded (Table 1). The final study group consisted of 65 subjects (43 exposed subjects and 22 referents). The exposed group and the referents were similar with regard to age, work schedule (i.e., shift work or daytime work), smoking habits, fish consumption, and amalgam fillings (Table 2). All subjects were examined at their ordinary workplace after having given written Table 1 The study group before and after exclusion of certain individuals
Total group Medication Beta-2 agonists Beta-blockers Serotonin inhibitors Neuroleptics
2 3 2 1
Diseases Diabetes Skull or whiplash injury Essential tremor
3 4 1
Out of form Lack of sleep Temporary nervousness Common cold
3 2 1
Ongoing pain in extremities Epicondylitis Shoulder pain Wrist pain
3 2 1
Final group, after exclusions
Exposed individuals
Referents
58
35
5
3
6
2
2
4
2
4
43
22
597
Table 2 Background characteristics of 43 exposed workers and 22 referents.
Mean age, years (range) Shift work, % (n) Smoking, % (n) All tobacco use, % (n) Amalgam fillings, % (n) Eating fish >2 times/week, % (n)
Exposed subjects (n = 43)
Referents (n = 22)
41 65% 26% 53% 86% 5%
40 68% 27% 32% 95% 14%
(25–65) (28) (11) (23) (37) (2)
(21–61) (15) (6) (7) (21) (3)
informed consent. The study was approved by the Ethics Committee of the Sahlgrenska University Hospital. 2.2. Clinical examinations The clinical examination included tests of coordination (diadochokinesis, and the finger–nose and knee–heel test), sensory function, deep tendon reflexes, gait, motor strength and tone, tremor, and Romberg’s test. All clinical tests were assessed as normal or abnormal, except for tremors (rest, postural, and intention tremors), which were graded as absent, slight (barely noticeable), or moderate (obvious, noticeable tremor, but <2 cm excursions). The Martin Vigorimeter was used to determine grip strength (Thorngren and Werner, 1979). Taken together, the questionnaires and clinical examination lasted for about 30 min. 2.3. Exposure assessment Data about exposure time (years) and work schedules were collected in the questionnaire, and current mercury exposure was assessed by determination of total mercury concentration in urine, corrected for creatinine (U-HgC). Based on data from previous concentrations of mercury in blood or urine (Sa¨llsten et al., 1990; and company records), a cumulative exposure index (U-Hgcum) was calculated for each exposed subject by summing up the mean yearly levels. Mean exposure for the past 5 years (U-Hgm5) was also calculated for each individual. Details of sampling and analytical procedures and assessment of previous mercury exposure (UHgcum) are given elsewhere (Wastensson et al., 2006). The mean exposure time for the 43 exposed workers was 15 years (median 13 years, range 2–32 years). As expected, the current mercury level was significantly higher in exposed workers (median 5.9 mg/g creatinine (mg/gC), mean 7.7 mg/gC, range 1.3– 25 mg/gC) than among referents (median 0.7 mg/gC, mean 0.9 mg/ gC, range 0.2–4.1 mg/gC). The median cumulative exposure index in the exposed group was 161 years mg/gC (mean 266 years mg/gC, range 8–1,440 years mg/gC), while the median value for U-Hgm5 was 6.8 mg/gC (n = 40, mean 10.9 mg/gC, range 2.1–37 mg/gC). A significant correlation was found between U-HgC and U-Hgm5 (rs = 0.77), but there were no significant correlations between current mercury exposure and either exposure time or UHgcum. Excluded individuals (n = 15 among exposed workers, n = 13 among referents) did not differ significantly in indices of mercury exposure from those included in the final group. 2.4. Quantitative assessment of neuromotor function Quantitative measurements of tremor, eye–hand coordination (using an EKM), and, rapid alternating movements (tested using the DIADO system) were performed after the clinical evaluation. The results from the quantitative assessment of tremor have been reported elsewhere (Wastensson et al., 2006). All quantitative measurements were conducted by the same neuropsychologist (DL), who was blinded with respect to the subjects’ exposure status.
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
598
Fig. 1. The eurythmokinesimeter.
Fig. 2. The diadochokinesimeter.
2.4.1. Quantitative assessment of rapid pointing movements (eurythmokinesimetry) The EKM system was developed to quantify eye–hand coordination in a pointing task (Beuter et al., 1999a). This system is composed of two targets, one distal and one proximal, each divided into three electrically isolated concentric areas, and a pointer. Before the evaluation began, the subjects were asked to sit down in front of the system and hold the pointer like a pen. The rod with the mounted targets was inclined at a 308 angle, and adjusted to such a distance that the subject’s arm was only slightly bent when the pointer was in contact with the distal target. The distance between the targets, from center to center, was kept at a fixed distance of 25 cm. In the evaluation procedure, the subjects were asked to alternately touch the center of each target, as precisely and quickly as possible (Fig. 1). The subjects were familiarized with the equipment before testing. Each recording lasted 30 s and was repeated twice with both hands alternating, with a 15 s pause between each recording. Finally, the recordings were transformed to nine calculated measures (Table 3), used to characterize the performance (Beuter et al., 1999a). For the statistical analyses, four values (two trials and two targets) were obtained for each characteristic and averaged to a mean for each hand. Where a value for a trial or target was missing, the average of the remaining values was calculated.
2.4.2. Quantitative assessment of alternating movements of the forearms (diadochokinesimetry) The diadochokinesimeter was developed in the 1990s to quantify the performance of rapid alternating movements of the forearms (Beuter et al., 1994, 1999b). The subjects were asked to sit down in front of the system and firmly hold a soft sphere in each hand (Fig. 2). The spheres were fixed to flexible rods mounted on another rod and connected to optical encoders by bendable joints. Before the beginning of the evaluation, the rod was adjusted so that the subject’s elbows were flexed at a 908 angle, and were free from any obstacles. The subjects were asked to hold the spheres so that the palms of their hands were facing each other, and then execute alternating movements of the forearms as fast as possible, with both hands moving simultaneously. All subjects were given time to get familiarized with the equipment before the recordings begun. Each recording lasted 5 s and was repeated twice, with a 15 s pause between each trial. To characterize the performance, six measures were calculated (Table 4), as described elsewhere (Beuter et al., 1999b). For the statistical analyses, the mean for each hand over the two trials for each characteristic was used. Where the value for one trial was missing, the remaining value was used in the calculations.
Table 3 Definitions of measures used to characterize rapid pointing movements, recorded with a eurythmokinesimeter Characteristicsa
Definitions
Speed
The number of times the target was struck (the number of events on target), divided by the sum of the times taken to reach the target before each event (in events per second). Larger scores indicate faster performance. The proportion of events involving a strike on target area A. Larger scores indicate a more precise performance. The proportion of strikes involving a strike on target area B, C, or D. Larger scores indicate a more imprecise performance. The average number of contacts per event. Smaller scores indicate lower disposition to sideslip across target areas or multiple contacts in one target area, and, therefore, better performance. The number of contacts, less the number of target areas contacted (averaged over events). Measurement of the number of extra contacts after the initial contact when there are multiple contacts on a target area. Smaller scores indicate less tremor interfering with the performance. The average duration of transportation of the hand from one target to another. Smaller scores indicate better performance. The average total duration of contacts on the target. Smaller scores indicate shorter contact and, therefore, better performance. The constant, k, is calculated as the average over events of k = t/log (2A/W), where t = the transit time to the target, A = the distance between the two target centers = 25 cm, and W = the approximate distance between the location of the contact(s) and the target center (for events with contacts on only one target area, the midpoint of the minimum and maximum distances of points in the area from the center was used; for events with contacts in two adjacent target areas, the distance of the separator between the two areas was used). This constant k should be a measure of inherent ability, independent of the subject’s choice in the speed/accuracy tradeoff. Smaller scores indicate better performance. The standard deviation of intervals between events. Events from both targets and trials were used together to calculate a single standard deviation. Smaller scores indicate more regular and, consequently, better performance.
Precision Imprecision Unsureness Tremor
Transit duration Contact duration Fitts’ law constant
Irregularity a
Definitions of characteristics from Beuter et al. (1999a).
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
599
Table 4 Definitions of measures used to characterize rapid alternating movements of the forearms, recorded with a diadochokinesimeter Characteristicsa
Definitions
Duration Range
Mean duration of an oscillation (in seconds). Larger values are considered worse in fast cadence conditions. Mean of total angular displacement (pronation and supination) per cycle (in degrees). Larger values indicate greater range, and, therefore, better performance. Mean velocity for each cycle, averaged over all cycles (in degrees per second). Larger values are better in fast cadence conditions. From the second differences of the signal, the sum of all absolute values of negative accelerations, where the velocity generally increases from zero to its maximum value, and all positive accelerations, where the velocity generally decreases from its maximum value to zero. The larger the number the more irregular the performance. The mean absolute value of velocity relative to peak velocity for ascending and descending segments, and averaged across all cycles. The larger the number the less difficulty there is at the turn. Maximum slope of a regression line, fitted to seven successive data points and averaged over each ascending and descending part of all the cycles (in degrees per sampling point; multiplied by 200 to obtain degrees per second). Larger values are better in fast cadence conditions.
Velocity Smoothness
Sharpness Maximum slope
a
Definitions of characteristics from Beuter et al. (1999b).
2.5. Statistics
3.2. Eurythmokinesimetry
For group comparisons, Student’s t-test or Wilcoxon’s rank sum test (n < 20, or skewed data) was used. Spearman’s correlation coefficients were used to evaluate associations between the outcome variables from the EKM and DIADO systems, mercury exposure, and potential confounders (age, shift work, smoking). Associations between the quantitative measurements of neuromotor function and mercury exposure were examined using multiple linear regression analysis, which allows for adjustment for potential confounders. The exposure indices (U-Hg, U-Hgm5, U-Hgcum) were included separately in the model. Variability between trials within individuals was assessed by the coefficient of variation (CV = relative standard deviation). p-Values of <0.05 (for twotailed tests) are reported as statistically significant. Statistics were calculated with the SAS statistical package, version 8.2 (SAS, 1999).
3.2.1. Associations with mercury exposure There were no significant group differences between the exposed subjects and referents in any of the characteristics used to describe eye–hand coordination, as measured by the EKM system (Table 6). In the multivariate analyses, unsureness was significantly associated with U-HgC and U-Hgm5 among exposed subjects, with age, work schedule, and smoking included in the model. In the dominant hand, these associations were, however, due to a single outlier. In the nondominant hand, these findings were supported by a similar significant association in the entire group, with age, work schedule and smoking included in the model. Tremor in both hands was significantly associated with UHgC and U-Hgm5 in the multivariate analyses, but in the dominant hand, all associations were due to the same outlier mentioned above. In the nondominant hand, no association between tremor and U-HgC was found in the entire group, and in fact, the exposed workers had less tremor compared with the referents (Table 6). After taking outliers into account, no significant associations were found between previous Hg exposure (U-Hgcum) and any of the outcome variables from the EKM test.
3. Results 3.1. Clinical examinations In 16 subjects, 12 exposed subjects and four referents, abnormalities, albeit mainly minor deviations, were found in the neurological tests. Of these, 10 subjects were classified as having abnormal tremor (rest, postural, or intention tremor). The results from the clinical examinations including excluded persons are summarized in Table 5.
3.2.2. Associations between measures, trials and targets Correlations between measures in the entire study group (n = 64, one subject missing) were calculated. As expected, we found that speed was highly negatively correlated to transit duration (dominant hand, rs = 0.84; nondominant hand, rs = 0.89) and precision (rs = 0.77; 0.80). Speed was also highly negatively correlated to Fitts’ law constant (dominant hand,
Table 5 Results from clinical examinations of 58 exposed subjects and 35 referents Final group (n = 65)
2
Grip strength, dominant hand, kp/cm (range) Grip strength, nondominant hand, kp/cm2 (range) Rest tremor, % (n) Postural tremor, % (n) Intention tremor (finger–nose), % (n) Abnormal knee–heel test, % (n) Dysdiadochokinesia, % (n) Abnormal Romberg’s test, % (n) Gait disturbance, % (n) Hyporeflexia, % (n) Hyperreflexia, % (n) Reduced sensation, % (n) Reduced sense of vibration, % (n)
Excluded persons (n = 28)
Exposed subjects (n = 43)
Referents (n = 22)
Exposed subjects (n = 15)
Referents (n = 13)
1.23 (0.75–1.60) 1.19 (0.80–1.60) 2% (1) 12% (5) 12% (5) (0) (0) (0) (0) 9% (4) (0) 5% (4) 5% (4)
1.18 (0.85–1.50) 1.13 (0.80–1.45) (0) 14% (3) 9% (2) (0) (0) (0) (0) (0) (0) 5% (1) 5% (1)
1.23 (1.00–1.70) 1.13 (0.70–1.50) (0) 27% (4) 33% (5) 7% (1) (0) (0) (0) 13% (2) (0) 13% (2) 27% (4)
1.23 (1.05–1.60) 1.15 (0.90–1.50) (0) 31% (4) 46% (6) (0) (0) 8% (1) (0) 8% (1) (0) 23% (3) 23% (3)
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
600
Table 6 Results of measurement of rapid pointing movements, using the eurythmokinesimeter§ Characteristicsa
p-Valueb
Group comparisons Exposed (n = 42)c
subjects
Correlation coefficients (Spearman’s)
Referents (n = 22)
All (n = 64)
Exposed subjects (n = 42)
U-HgC
U-HgC
S.D.
Mean
S.D.
Dominant hand Speed (mm/s) Precision Imprecision Unsureness Tremor Transit duration Contact duration Fitts’ law constant Irregularity
1.10 0.48 0.69 1.38 0.22 0.77 0.24 0.14 0.15
0.33 0.20 0.17 0.34 0.21 0.19 0.20 0.03 0.10
1.06 0.55 0.68 1.46 0.24 0.78 0.23 0.14 0.13
0.25 0.23 0.20 0.27 0.15 0.19 0.16 0.03 0.07
0.62 0.19 0.74 0.34 0.78 0.85 0.87 0.85 0.51
0.06 0.02 0.07 0.07 0.03 0.01 0.04 0.02 0.07
0.06 0.18 0.11 0.10 0.15 0.02 0.09 0.01 0.04
0.15 0.15 0.16 0.15 0.10 0.22 0.01 0.23 0.08
0.05 0.03 0.07 0.04 0.02 0.05 0.01 0.07 0.23
Nondominant hand Speed (mm/s) Precision Imprecision Unsureness Tremor Transit duration Contact duration Fitts’ law constant Irregularity
1.05 0.47 0.78 1.52 0.24 0.81 0.24 0.15 0.14
0.34 0.19 0.13 0.30 0.15 0.20 0.19 0.03 0.07
1.00 0.51 0.74 1.56 0.32 0.82 0.26 0.15 0.14
0.27 0.25 0.17 0.40 0.21 0.23 0.19 0.03 0.07
0.62 0.42 0.35 0.64 0.13 0.89 0.69 0.92 0.97
0.04 0.01 0.12 0.09 0.09 0.03 0.00 0.05 0.05
0.03 0.03 0.17 0.09 0.06 0.04 0.04 0.03 0.05
0.16 0.08 0.26 0.03 0.01 0.17 0.07 0.18 0.02
0.01 0.00 0.15 0.22 0.12 0.04 0.03 0.05 0.12
§
U-Hgcum
U-Hgm5d
Mean
For an explanation of the characteristics, see Table 3. a Mean of four recordings. b t-test, adjusted for unequal variance. c One subject missing. d n = 39.
rs = 0.81; nondominant hand, rs = 0.86), a measure which is independent of the subject’s choice of speed/accuracy strategy. Furthermore, strong associations were found between unsureness and tremor (dominant hand, rs = 0.85; nondominant hand, rs = 0.76). Intraindividual variability for the four values (two trials and two targets) for each hand was calculated for all characteristics in the entire group (n = 64). It was low for most characteristics, with the CV ranging from 9% to 31%. However, the variability in tremor was high in both hands (dominant hand, 70%; nondominant hand, 83%), and for irregularity, the CV was 58% in the nondominant hand and >100% in the dominant hand. 3.2.3. Associations with age, shift work, and smoking habits In the entire study group (n = 64, one subject missing), increasing age was associated with a decrease in speed (dominant hand, rs = 0.33, p = 0.007; nondominant hand rs = 0.30, p = 0.02), and with longer contact duration between the stylus and metal targets (dominant hand, rs = 0.42, p = 0.0005; nondominant hand, rs = 0.44, p = 0.0003). In the nondominant hand, older subjects had more tremor (rs = 0.29, p = 0.02), but in contrast, a more precise performance (rs = 0.29, p = 0.02). Smokers had significantly more tremor ( p = 0.048) and a tendency to sideslip across target areas ( p = 0.02) in the dominant hand compared with nonsmokers. No significant differences were found between shift workers and daytime workers in any of the outcome variables from the EKM test. 3.3. Diadochokinesimetry 3.3.1. Associations with mercury exposure There were no significant overall group differences between exposed subjects and referents in any of the outcome variables from the test of rapid alternating movements of the forearms (Table 7). There were, however, some inverse associations
between Hg exposure, on the one hand, and duration, velocity, sharpness, and range, on the other (Table 7). In the multivariate analyses, an inverse association was found between velocity in the dominant hand and Hg exposure (U-HgC and U-Hgm5) among exposed subjects, and it remained statistically significant with age and work schedule included in the model. Furthermore, the ability to make fast turns (sharpness) was inversely associated with U-HgC and U-Hgm5 among exposed subjects in the dominant hand, and was close to significant with age and shift work included in the model. A further analysis comparing a subgroup of the exposed workers with the highest U-HgC (n = 22, U-HgC 5.9; median level) or subjects with the highest U-Hgm5 (n = 20, U-Hgm5 6.75; median level) with the referents, did not reveal any differences with regard to velocity or sharpness. Finally, range in the dominant hand was inversely associated with U-HgC and U-Hgm5, but these associations disappeared when a single outlier was removed. No significant associations were found between previous Hg exposure and the outcome variables from the DIADO test in the multivariate analyses. 3.3.2. Associations between measures and trials In the entire study group (n = 65), duration was highly negatively correlated to velocity (dominant hand, rs = 0.73; nondominant hand, rs = 0.70) and sharpness (rs = 0.76; 0.79). Moderate to strong associations were found between maximum slope and range (dominant hand, rs = 0.54; nondominant hand, rs = 0.60), and velocity (rs = 0.67; 0.68). For most characteristics, correlations between the first and second trials were high in both hands. Expressed as the CV, the intraindividual variation for duration, range, velocity, sharpness, and maximum slope was about 8% (range 6–12%). The variability in smoothness was, however, very high, and even when an outlier was removed the CV was >100% in both hands.
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
601
Table 7 Results of measurement of rapid alternating movements, using a diadochokinesimeter§ Characteristicsa
Exposed (n = 43)
Dominant hand Duration Range Velocity Smoothness (10 Sharpness Maximum slope
2
)
Nondominant hand Duration Range Velocity Smoothness (10 2) Sharpness Maximum slope §
p-Valuec
Group comparisons subjects
Correlation coefficientsb (Spearman’s)
Referents (n = 22)
All (n = 65)
Exposed subjects (n = 43)
U-HgC
U-HgC
U-Hgcum
U-Hgm5d
Mean
S.D.
Mean
S.D.
4.76 333.76 75.93 2.81 0.49 7.61
1.45 54.34 20.06 0.13 0.08 1.42
5.22 347.77 74.29 0.75 0.49 7.50
2.10 63.99 22.26 0.54 0.10 1.70
0.37 0.36 0.77 0.30 0.98 0.77
0.05 0.25* 0.19 0.01 0.19 0.02
0.23 0.24 0.40* 0.06 0.29 0.12
0.35* 0.19 0.24 0.03 0.21 0.11
0.26 0.11 0.29 0.00 0.26 0.06
4.78 333.06 74.85 0.58 0.50 7.31
1.44 55.99 18.38 0.45 0.08 1.36
5.20 335.03 71.22 0.77 0.50 7.05
1.99 54.79 19.58 0.93 0.09 1.59
0.33 0.89 0.46 0.37 0.77 0.50
0.07 0.04 0.08 0.15 0.09 0.07
0.25 0.03 0.20 0.01 0.19 0.00
0.37* 0.22 0.23 0.17 0.32* 0.00
0.26 0.03 0.08 0.19 0.18 0.08
For an explanation of the characteristics, see Table 4. a Mean of two recordings. b *p-value < 0.05. c t-test, adjusted for unequal variance. d n = 40.
3.3.3. Associations with age, shift work, and smoking habits A high age was associated with reduced sharpness (rs = 0.32, p = 0.008) in the nondominant hand (n = 65), but the association did not reach statistical significance in the dominant hand. Shift workers performed alternating movements significantly faster in both hands compared with daytime workers. Nonsmokers had a significantly greater range and, consequently, better performance than smokers in both hands. 3.4. Comparison between the clinical assessment and the quantitative tests The original group (n = 93) was used for comparing clinical findings with the outcome variables in the quantitative tests. First, all subjects with neurological signs (n = 31) were compared with the remaining subjects (n = 62). Subjects with neurological abnormalities were significantly older (mean age 47 years, compared with 39 years among the other subjects) but did not differ in any of the measurements in the EKM or DIADO tests. Interestingly, subjects with abnormal tremor (rest, postural or intention tremors) (n = 21) had significantly shorter duration, higher velocity, and maximum slope (better performance) in both hands than other subjects in the DIADO test. The difference remained significant for velocity and maximum slope when subjects with only postural (n = 15) or intention tremor (n = 18) were compared with the other subjects. No significant differences were shown for the EKM test. 4. Discussion 4.1. Discussion of findings Overall, this study does not indicate any significant adverse effects of long-term low level Hg exposure on certain measures of neuromotor function. The clinical examinations did not reveal more neurological abnormalities among exposed subjects than the referent group. Some findings in the EKM test, however, indicate that current mercury exposure may have an effect on eye–hand coordination (unsureness, tremor) in the nondominant hand. In
the DIADO test, no significant associations with Hg exposure were found, except for a possible effect on velocity in the dominant hand among exposed subjects only. A tendency towards more neurological abnormalities has been reported in studies of workers with high current Hg exposure, down to U-Hg levels of around 70 mg/gC (Miller et al., 1975; Ehrenberg et al., 1991; Urban et al., 1996). The abnormal findings reported include eyelid fasciculations (Miller et al., 1975), static tremor (Ehrenberg et al., 1991; Urban et al., 1996), hyperactive tendon reflexes (Miller et al., 1975), absence of ankle jerk (Urban et al., 1996), dysdiadochokinesis, difficulties with heel–toe gait, and rest and intention tremors (Ehrenberg et al., 1991). In accordance with the present study, no increase in neurological abnormalities has been shown among exposed workers at lower (U-Hg 20–25 mg/gC or mg/L) exposure levels (Chapman et al., 1990; Langworth et al., 1992). Effects on motor speed and coordination have been reported among workers with high or moderate Hg exposure (>U-Hg 25 mg/ gC or mg/L), with several studies reporting reduced performance on finger tapping (Miller et al., 1975; Langolf et al., 1978; Gu¨nther et al., 1996) and eye–hand coordination (Miller et al., 1975; Langolf et al., 1978; Roels et al., 1982, 1989; Williamson et al., 1982; Piikivi et al., 1984; Gu¨nther et al., 1996). The results from studies with lower exposure levels (U-Hg 25 mg/g or mg/L) are somewhat contradictory. In a study of 89 chloralkali workers with an average U-Hg of around 25 mg/gC (Langworth et al., 1992), no differences in finger tapping or eye– hand coordination were found between exposed subjects and referents. The performance in these tests was, however, negatively correlated to the relative number of U-Hg peaks >30 mg/L among exposed workers. In another study of chloralkali workers, with a mean U-Hg of 18 mg/g, no effect on eye–hand coordination or finger tapping was found, and surprisingly, the exposed workers performed better (Piikivi and Ha¨nninen, 1989). By contrast, reduced performance in finger tapping was shown among exposed subjects compared with the referent group in a study of 88 lamp factory workers with similar exposure levels as in the studies above (Liang et al., 1993).
602
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
A possible explanation for the discrepancy in observed effects among currently exposed workers may be different historical Hg exposure in the study populations. Several studies show an excess of abnormal clinical findings among previously exposed workers even long after cessation of exposure (Albers et al., 1988; Andersen et al., 1993; Powell, 2000; Letz et al., 2000; Frumkin et al., 2001). In addition, effects on finger tapping (Frumkin et al., 2001; Kishi et al., 1994), eye–hand coordination and manual dexterity (Frumkin et al., 2001; Kishi et al., 1994; Mathiesen et al., 1999; Powell, 2000) were shown. In these studies, however, the historical exposure levels were generally high, and a recent study on previously exposed workers with similar cumulative Hg exposure as for the workers in the present study showed no effect on motor speed and coordination, which is in accordance with our findings (BastPettersen et al., 2005). An association between reduced motor speed, as evaluated with finger tapping, and low exposure levels (U-Hg 10 mg/g) has been shown in a study of industrial workers including workers from the chloralkali industry (Lucchini et al., 2002) and in two studies on dental personnel (Ngim et al., 1992; Echeverria et al., 1998). These findings are, however, in contrast to the absence of such associations in another study of chloralkali workers with low exposure levels (Ellingsen et al., 2001). Our findings of possible effects on eye–hand coordination (evaluated using the EKM test) have little support from negative studies using other tests of manual dexterity, such as the grooved pegboard test (Ngim et al., 1992; Ellingsen et al., 2001) and the one-hole pins test (Echeverria et al., 1998). Impaired performance of rapid alternating movements has not been described at lower exposure levels, and as far as we know, quantitative measurement of these movements has not previously been applied in subjects exposed to Hg0. Consequently, our finding of a possible effect on velocity in the dominant hand must be confirmed by other studies. As expected, older subjects were slower, mainly due to longer contact with the targets, and had more tremor (nondominant hand) in the hand–eye coordination test (using the EKM system). We could not, however, confirm an association between age and more irregularity or high values on Fitts’ law constant, indicating a general impairment of performance by age, as described in another study (Beuter et al., 1999). Except for the reduced ability, among older subjects, to rapidly stop and start alternating movements (sharpness) in the nondominant hand, age was not significantly associated with any of the other outcome variables in the DIADO test. Possible explanations for the modest effect of age on these quantitative tests in contrast to other studies (Beuter et al., 1999) may be that our subjects were a selected group (i.e., they were healthier and all still of working age), or that the sample size was small. Shift work was not found to impair the ability to execute tasks demanding good eye–hand coordination or the performance of rapid alternating movements, which could have been expected. Our finding that shift workers were faster than daytime workers in the DIADO test could possibly be a ‘‘healthy workers effect’’ (i.e., shift workers may tend to move to daytime work when they get health problems). The fact that the shift workers had fewer abnormal findings in the clinical examinations (21% vs. 32% among daytime workers) speaks in favor of this explanation. In the EKM test, smokers had more tremor and tendency to sideslip across target areas compared with nonsmokers. In accordance with our findings, increased tremor related to smoking habits has been reported in other occupational studies using quantitative tests of hand steadiness and tremor (Ellingsen et al., 2001, 2006; Bast-Pettersen et al., 2004), and has been shown to affect a variety of hand activities (Louis, 2007). In a study of 297 subjects from the general population in Quebec, Canada, with low
level exposure to manganese (Beuter et al., 1999), no associations between smoking habits and any of the outcome variables from the DIADO test were shown. Consequently, our finding of a possible effect of smoking habits on range in the DIADO test has to be confirmed by other studies. 4.2. Aspects of validity The exposed subjects and the referents were similar in important background characteristics, such as age, smoking habits, and work schedule, indicating good comparability between groups. Other sources that may give a small contribution to Hg0 exposure, i.e., amalgam fillings or high fish consumption were similar among exposed chloralkali workers and referents. All workers were employed at the same plants, and the participation rate was high, indicating no major selection bias. We chose to exclude all subjects with diseases or medication that may affect neuromotor function, in order to reduce ‘‘noise’’ and to be able to detect more subtle effects that might be caused by Hg exposure. On the other hand, when the number of subjects was reduced, only few subjects with high U-Hg levels were left among our exposed subjects. Consequently, if there is a threshold, and if Hg exposure affects only subjects with the highest U-Hg levels, such an effect could not be detected. In fact, the ‘outlier’ mentioned above, with abnormal results in some EKM measures (unsureness, tremor), had experienced relatively high exposure in the last 5 years (mean 37 mg/gC). Since no other plausible explanation (family history of tremor, disease or medication) was found, Hg exposure may be considered as a possible cause to the abnormal results in this subject. In the present study, highly sensitive methods were used to evaluate possible effects of low level mercury exposure on neuromotor function. To be applicable in the occupational field, these methods must be reliable and simple to use, however. For the EKM test, most characteristics were stable over trials; however, two trials may be insufficient for tremor and irregularity, as these characteristics show high intraindividual variation. In the DIADO test, the recordings were made in fast cadence, with both hands moving simultaneously. This condition is anticipated to exacerbate any subclinical adverse effects on the CNS more than other conditions (Beuter et al., 1994); also, the characteristics used in our study have been shown to be consistent between trials in fast cadence (Beuter et al., 1999b). However, our results indicate that the number of trials (or duration of test) is insufficient for smoothness, which has high intraindividual variability. Our study group (n = 43 + 22) was small, but not extremely small. We performed power calculations on several of the outcome variables and found that the statistical power was moderate to good. For example, in the present study, the power was 80% (a = 0.05) to detect a 10% difference in mean Fitts’ law constant (EKM test) and sharpness (DIADO test), or a 20% difference in mean precision (EKM test) between the groups. The EKM system is anticipated to be similar to the finger–nose test. Surprisingly, the comparison between the clinical evaluation (intention tremor) and the quantitative measures from the EKM test showed no associations but is in accordance with another study (Beuter et al., 1999a). In the finger–nose test, the subject is asked to alternately touch the end of the nose with the tip of the index finger; and the examiner classifies the performance as normal or abnormal (slight or obvious tremor). As has been indicated by Notermans et al. (1994), these clinical tests are usually not sufficiently sensitive to detect subclinical changes in neuromotor function, and contain a considerable degree of subjectivity when criteria are applied (McKeown-Eyssen et al., 1990). In addition, marked interobserver variation has been
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
reported in the assessment of signs of neurological abnormalities (McKeown-Eyssen et al., 1990). In the present study, two physicians conducted the clinical examinations, and consequently, variability between observers may have affected the results. Moreover, the quantitative test measures several distinct aspects of the performance, and is, in contrast to the clinical evaluation, a continuous variable. There was also a certain time delay (about 15– 30 min) between the two tests. Finally, the two tests may not be completely comparable. In the finger–nose test, the subjects performed the test with their eyes closed, while the EKM test demands good function of the visual pathways. In the present study, all subjects were classified as normal in the clinical assessment of diadochokinesis; therefore, a comparison with the quantitative method (DIADO test) was not possible. We have no obvious explanation for our findings that subjects with tremor at clinical examination performed better in the DIADO test compared with subjects without clinical tremor. In conclusion, the present study reveals no significant adverse effects of exposure to Hg0 on certain aspects (eye–hand coordination, diadochokinesis) of neuromotor function at low exposure levels. The lack of agreement between clinical tests (finger–nose test) and quantitative assessment (EKM test) was surprising. In general, the quantitative systems used in this study were reliable and easy to handle. They should be suitable for studies of groups exposed to neurotoxic chemicals. However, certain modifications (duration of test, number of trials) may be useful to compensate for the high intraindividual variability in some outcome measures. Acknowledgements The authors would like to thank the participants in this study and the staff at the local Occupational Health service units for their assistance. We are grateful to Annika Claesson, for helpful assistance with collecting and handling the urine samples, and the late Andrejs Schu¨tz, head of the laboratory at the Department of Occupational and Environmental Medicine in Lund. Finally, we wish to thank Eva Andersson for statistical advice.
References Albers JW, Kallenbach LR, Fine LJ, Langolf GD, Wolfe RA, Donofrio PD, et al. Neurological abnormalities associated with remote occupational elemental mercury exposure. Ann Neurol 1988;24:651–9. Andersen A, Ellingsen DG, Mørland T, Kjuus H. A neurological and neurophysiological study of chloralkali workers previously exposed to mercury vapour. Acta Neurol Scand 1993;88:427–33. Bast-Pettersen R, Ellingsen DG, Hetland SM, Thomassen Y. Neuropsychological function in manganese alloy plant workers. Int Arch Occup Environ Health 2004;77:277–87. Bast-Pettersen R, Ellingsen DG, Efskind J, Jordskogen R, Thomassen Y. A neurobehaviorial study of chloralkali workers after the cessation of exposure to mercury vapor. Neurotoxicology 2005;26:427–37. Beuter A, Mergler D, De Geoffroy A, Carrie`re L, Be´langer S, Varghese L, et al. Diadochokinesimetry: a study of patients with Parkinson’s disease and manganese exposed workers. Neurotoxicology 1994;15:655–64. Beuter A, Edwards R, De Geoffroy A, Mergler D, Hudnell K. Quantification of neuromotor function for detection of the effects of manganese. Neurotoxicology 1999;20:355–66. Beuter A, De Geoffroy A, Edwards R. Quantitative analysis of rapid pointing movements in Cree subjects exposed to methylmercury and in subjects with neurological deficits. Environ Res 1999a;80:50–63. Beuter A, De Geoffroy A, Edwards R. Analysis of rapid alternating movements in Cree subjects exposed to methylmercury and in subjects with neurological deficits. Environ Res 1999b;80:64–79.
603
Chapman LJ, Sauter SL, Henning RA, Dodson VN, Reddan WG, Matthews CG. Differences in frequency of finger tremor in otherwise asymptomatic mercury workers. Br J Ind Med 1990;47:838–43. Clarkson TW, Magos L. The toxicology of mercury and its chemical compounds. Crit Rew Toxicol 2006;36:609–62. Echeverria D, Aposhian HV, Woods JS, Heyer NJ, Aposhian MM, Bittner AC Jr, et al. Neurobehavioral effects from exposure to dental amalgam Hg0: new distinctions between recent exposure and Hg body burden. FASEB J 1998;12:971–80. Ehrenberg RL, Vogt RL, Smith AB, Brondum J, Brightwell WS, Hudson PJ, et al. Effects of elemental mercury exposure at a thermometer plant. Am J Ind Med 1991;19:495–507. Ellingsen DG, Bast-Pettersen R, Efskind J, Thomassen Y. Neuropsychological effects of low mercury vapor exposure in chloralkali workers. Neurotoxicology 2001;22:249–58. Ellingsen DG, Bast-Pettersen R, Efskind J, Gjølstad M, Olsen R, Thomassen Y, et al. Hand tremor related to smoking habits and the consumption of caffeine in male industrial workers. Neurotoxicology 2006;27:525–33. Fawer RF, De Ribaupierre Y, Guillemin MP, Berode M, Lob M. Measurement of hand tremor induced by industrial exposure to metallic mercury. Br J Ind Med 1983;40:204–8. Fimbel EJ, Domingo PP, Lamoureux D, Beuter A. Automatic detection of movement disorders using recordings of rapid alternating movements. J Neurosci Methods 2005;146:183–90. Frumkin H, Letz R, Williams PL, Gerr F, Pierce M, Sanders A, et al. Health effects of long-term mercury exposure among chloralkali plant workers. Am J Ind Med 2001;39:1–18. Gu¨nther W, Sietman B, Seeber A. Repeated neurobehavioral investigations in workers exposed to mercury in a chloralkali plant. Neurotoxicology 1996;17:605–14. Kishi R, Doi R, Fukuchi Y, Satoh H, Satoh T, Ono A, et al. Residual neurobehavioral effects associated with chronic exposure to mercury vapour. Occup Environ Med 1994;51:35–41. Langolf GD, Chaffin DB, Henderson R, Whittle HP. Evaluation of workers exposed to elemental mercury using quantitative tests of tremor and neuromuscular function. Am Ind Hyg Assoc J 1978;39:976–84. Langworth S, Almkvist O, So¨derman E, Wikstro¨m BO. Effects of occupational exposure to mercury vapour on the central nervous system. Br J Ind Med 1992;49:545–55. Letz R, Gerr F, Cragle D, Green RC, Watkins J, Fidler AT. Residual neurologic deficits 30 years after occupational exposure to elemental mercury. Neurotoxicology 2000;21:459–74. Liang YX, Sun RK, Sun Y, Chen ZQ, Li LH. Psychological effects of low exposure to mercury vapor: application of a computer-administered neurobehavioral evaluation system. Environ Res 1993;60:320–7. Louis ED. Kinetic tremor: differences between smokers and nonsmokers. Neurotoxicology 2007;28:569–75. Lucchini R, Cortesi I, Facco P, Benedetti L, Camerino D, Carta P, et al. Neurotoxic effect of exposure to low doses of mercury. Med Lav 2002;93:202–14. Mathiesen T, Ellingsen DG, Kjuus H. Neuropsychological effects associated with exposure to mercury vapor among former chloralkali workers. Scand J Work Environ Health 1999;25:342–50. McKeown-Eyssen GE, Ruedy J, Hogg S, Read Guernsey J, Woods I. Validity and reproducibility of a screening examination for neurological abnormality in persons exposed to methylmercury. J Clin Epidemiol 1990;43:489–98. Meyer-Baron M, Schaeper M, Van Thriel C, Seeber A. Neurobehavioral test results and exposure to inorganic mercury: in search of dose-response relations. Arch Toxicol 2004;78:207–11. Miller JM, Chaffin DB, Smith RG. Subclinical psychomotor and neuromuscular changes in workers exposed to inorganic mercury. Am Ind Hyg Assoc J 1975;36:725–33. Ngim CH, Foo SC, Boey KW, Jeyaratnam J. Chronic neurobehavioural effects of elemental mercury in dentists. Br J Ind Med 1992;49:782–90. Notermans NC, Van Dijk GW, Van der Graaf Y, Van Gijn J, Wokke JHJ. Measuring ataxia: quantification based on the standard neurological examination. J Neurol Neurosurg Psychiatry 1994;57:22–6. Piikivi L, Ha¨nninen H. Subjective symptoms and psychological performance of chlorine-alkali workers. Scand J Work Environ Health 1989;15:69–74.
604
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
Piikivi L, Ha¨nninen H, Martelin T, Mantere P. Psychological performance and long-term exposure to mercury vapors. Scand J Work Environ Health 1984;10:35–41. Powell TJ. Chronic neurobehavioral effects of mercury poisoning on a group of Zulu chemical workers. Brain Inj 2000;14:797–814. Roels H, Lauwerys R, Buchet JP, Bernard A, Barthels A, Oversteyns M, et al. Comparison of renal function and psychomotor performance in workers exposed to elementary mercury. Int Arch Occup Environ Health 1982;50:77–93. Roels H, Abdeladim S, Braun M, Malchaire J, Lauwerys R. Detection of hand tremor in workers exposed to mercury vapor: a comparative study of three methods. Environ Res 1989;49:152–65. Rohling ML, Demakis GJ. A meta-analysis of the neuropsychological effects of occupational exposure to mercury. Clin Neuropsychol 2006;20:108–32. Sa¨llsten G, Barrega˚rd L, Ja¨rvholm B. Mercury in the Swedish chloralkali industry—an evaluation of the exposure and preventive measures over 40 years. Ann Occup Hyg 1990;34:205–14.
SAS Institute, Inc. Statistical software version 8. SAS Institute, Inc., Cary, NC, USA, 1999. Thorngren KG, Werner CO. Normal grip strength. Acta Orthop Scand 1979;50:255–9. Urban P, Lukas E, Benicky L, Moscovicova E. Neurological and electrophysiological examination of workers exposed to mercury vapors. Neurotoxicology 1996;17:191–6. Wastensson G, Lamoureux D, Sa¨llsten G, Beuter A, Barrega˚rd L. Quantitative tremor assessment in workers with current low exposure to mercury vapor. Neurotoxicol Teratol 2006;28:681–93. WHO (World Health Organization). Concise international chemical assessment document 50. Elemental mercury and inorganic mercury compounds: human health aspects. WHO, Geneva, Switzerland, 2003. Williamson AM, Teo RKC, Sanderson J. Occupational mercury exposure and its consequences for behaviour. Int Arch Environ Health 1982;50: 273–86.
Contents lists available at ScienceDirect
NeuroToxicology
Quantitative assessment of neuromotor function in workers with current low exposure to mercury vapor Gunilla Wastensson a,*, Daniel Lamoureux b, Gerd Sa¨llsten a, Anne Beuter b,c, Lars Barrega˚rd a a
Department of Occupational and Environmental Medicine, Sahlgrenska University Hospital, and University of Go¨teborg, Box 414, SE-405 30 Go¨teborg, Sweden Centre de Neuroscience de la Cognition, Universite´ du Que´bec a` Montre´al, Montre´al, Que´bec, C.P. 8888, Canada H3C 3P8 c Institut de Cognitique, Universite´ Victor Segalen Bordeaux 2, Zone Nord, 146 rue Leo Saignat, 33076 Bordeaux Cedex, France b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 14 November 2007 Accepted 5 March 2008 Available online 20 March 2008
Evaluation of neuromotor function has been used in several epidemiological studies of workers with long-term exposure to mercury vapor (Hg0). Some recent studies indicate adverse effects at relatively low exposure levels. In the present study, we used sensitive quantitative methods, developed specifically to detect subtle effects of exposure to toxins on motor function. After exclusion of individuals with neurological diseases or other conditions that may affect performance, 43 chloralkali workers with current low exposure to Hg0, and 22 age-matched referents remained for further analysis. The median urinary mercury concentration in exposed workers was 5.9 mg/g (range 1.3–25) creatinine (mg/gC), while that in referents was 0.7 mg/gC (range 0.2–4.1). The mean exposure time was 15 years, and the median cumulative mercury index was 161 years mg/gC in exposed workers. A eurythmokinesimeter (EKM) was used to quantify eye–hand coordination, and a diadochokinesimeter, to measure rapid alternating rotation of the forearms. In general, the differences in performance between the exposed workers and the referents were small. Age was associated with a decrease in speed, more tremor, and longer contact duration between the stylus and the metal targets in performance of rapid pointing movements. Smokers had significantly more tremor, and more contacts per event in the EKM test, than nonsmokers. Taking age, shift work, and smoking habits into account, no significant associations with current or cumulative mercury exposure were found for the majority of the outcome variables from the quantitative tests. In general, this study indicates no significant adverse effects of Hg0 on neuromotor function at the exposure levels studied. ß 2008 Elsevier Inc. All rights reserved.
Keywords: Neuromotor function Mercury vapor Quantitative methods
1. Introduction Exposure to mercury vapor (Hg0) may cause adverse effects on many organs, including the central nervous system (CNS). The major routes of human exposure are inhalation of Hg0 from occupational sources, and dental amalgam (Clarkson and Magos, 2006). About 80% of the inhaled vapor is absorbed through the lungs. The highly lipophilic vapor spreads throughout the body dissolved in the blood, and readily crosses the blood–brain barrier (WHO, 2003). Inside the cells, Hg0 is oxidized to divalent mercury (Hg2+), which is assumed to be the proximate toxic agent (Clarkson and Magos, 2006). After exposure, most of the mercury in the brain is cleared with a short half-time, but a fraction may have a much longer half-time, of several years (WHO, 2003). Excretion is via urine and feces, with a whole-body, half-time of about 60 days.
* Corresponding author. Tel.: +46 31 786 2894; fax: +46 31 40 97 28. E-mail address: [email protected] (G. Wastensson). 0161-813X/$ – see front matter ß 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2008.03.005
Measurement of mercury in urine is considered to be the best indicator of body burden of mercury with long-term exposure, and is commonly used as biomarker of exposure (WHO, 2003). There is a wide range of cognitive, personality, sensory, and motor disturbances reported after exposure to Hg0 including tremor, emotional lability, polyneuropathy, and impaired performance in tests of cognitive function and psychomotor skills (WHO, 2003). Even if little is known about the exact pattern of distribution in the CNS among humans (Clarkson and Magos, 2006), the extent and variety in neuropsychological impairment following Hg0 exposure imply that most structures in the CNS are affected. One of the earliest sign of mercury intoxication is a typical intentional tremor, indicating impairment of the cerebellum, a structure involved in coordination and voluntary movements. Traditionally, the finger–nose test was used in periodic examinations of mercury-exposed workers; however, decreasing exposure levels due to improved hygienic conditions require more precise and sensitive tools. Some studies using sensitive tests show a possible effect on psychomotor skills, such as reduced motor speed
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
(Ngim et al., 1992; Echeverria et al., 1998; Lucchini et al., 2002), and tremor (Chapman et al., 1990; Fawer et al., 1983; Langworth et al., 1992) even at relatively low exposure levels. Moreover, meta-analyses performed on studies of mercury-exposed workers have shown a larger impairment on motor performance compared with other neuropsychological domains such as memory and attention (Meyer-Baron et al., 2004; Rohling and Demakis, 2006). The aim of this study was to investigate effects on neuromotor function in chloralkali workers with current low level exposure to Hg0 using new sensitive quantitative methods. The eurythmokinesimeter (EKM) used in the present study quantifies rapid and precise proximo-distal movements in a pointing task supposed to be similar to the finger–nose test. The ability to perform rapid alternating movements (pronation/supination of the forearms), diadochokinesis, is quantitatively measured by the diadochokinesimeter (DIADO). The reliability and the sensitivity of these techniques have been evaluated in populations exposed to methylmercury (Beuter et al., 1999a, 1999b) and manganese (Beuter et al., 1994; Beuter et al., 1999), and among subjects with Parkinson’s disease and other neurological deficits (Beuter et al., 1994, 1999a, 1999b; Fimbel et al., 2005).
2. Methods 2.1. Subjects We recruited 58 mercury-exposed workers and 35 unexposed referents, all men, from two different chloralkali plants located in the same geographic area of the southwest region of Sweden. The selection procedure has been described elsewhere in detail (Wastensson et al., 2006). Information concerning relevant background variables, and the subjects’ state of health was collected in a questionnaire and by an additional interview before a physical examination was performed. Individuals with diseases, medication, or other circumstances that could possibly affect neuromotor function were excluded (Table 1). The final study group consisted of 65 subjects (43 exposed subjects and 22 referents). The exposed group and the referents were similar with regard to age, work schedule (i.e., shift work or daytime work), smoking habits, fish consumption, and amalgam fillings (Table 2). All subjects were examined at their ordinary workplace after having given written Table 1 The study group before and after exclusion of certain individuals
Total group Medication Beta-2 agonists Beta-blockers Serotonin inhibitors Neuroleptics
2 3 2 1
Diseases Diabetes Skull or whiplash injury Essential tremor
3 4 1
Out of form Lack of sleep Temporary nervousness Common cold
3 2 1
Ongoing pain in extremities Epicondylitis Shoulder pain Wrist pain
3 2 1
Final group, after exclusions
Exposed individuals
Referents
58
35
5
3
6
2
2
4
2
4
43
22
597
Table 2 Background characteristics of 43 exposed workers and 22 referents.
Mean age, years (range) Shift work, % (n) Smoking, % (n) All tobacco use, % (n) Amalgam fillings, % (n) Eating fish >2 times/week, % (n)
Exposed subjects (n = 43)
Referents (n = 22)
41 65% 26% 53% 86% 5%
40 68% 27% 32% 95% 14%
(25–65) (28) (11) (23) (37) (2)
(21–61) (15) (6) (7) (21) (3)
informed consent. The study was approved by the Ethics Committee of the Sahlgrenska University Hospital. 2.2. Clinical examinations The clinical examination included tests of coordination (diadochokinesis, and the finger–nose and knee–heel test), sensory function, deep tendon reflexes, gait, motor strength and tone, tremor, and Romberg’s test. All clinical tests were assessed as normal or abnormal, except for tremors (rest, postural, and intention tremors), which were graded as absent, slight (barely noticeable), or moderate (obvious, noticeable tremor, but <2 cm excursions). The Martin Vigorimeter was used to determine grip strength (Thorngren and Werner, 1979). Taken together, the questionnaires and clinical examination lasted for about 30 min. 2.3. Exposure assessment Data about exposure time (years) and work schedules were collected in the questionnaire, and current mercury exposure was assessed by determination of total mercury concentration in urine, corrected for creatinine (U-HgC). Based on data from previous concentrations of mercury in blood or urine (Sa¨llsten et al., 1990; and company records), a cumulative exposure index (U-Hgcum) was calculated for each exposed subject by summing up the mean yearly levels. Mean exposure for the past 5 years (U-Hgm5) was also calculated for each individual. Details of sampling and analytical procedures and assessment of previous mercury exposure (UHgcum) are given elsewhere (Wastensson et al., 2006). The mean exposure time for the 43 exposed workers was 15 years (median 13 years, range 2–32 years). As expected, the current mercury level was significantly higher in exposed workers (median 5.9 mg/g creatinine (mg/gC), mean 7.7 mg/gC, range 1.3– 25 mg/gC) than among referents (median 0.7 mg/gC, mean 0.9 mg/ gC, range 0.2–4.1 mg/gC). The median cumulative exposure index in the exposed group was 161 years mg/gC (mean 266 years mg/gC, range 8–1,440 years mg/gC), while the median value for U-Hgm5 was 6.8 mg/gC (n = 40, mean 10.9 mg/gC, range 2.1–37 mg/gC). A significant correlation was found between U-HgC and U-Hgm5 (rs = 0.77), but there were no significant correlations between current mercury exposure and either exposure time or UHgcum. Excluded individuals (n = 15 among exposed workers, n = 13 among referents) did not differ significantly in indices of mercury exposure from those included in the final group. 2.4. Quantitative assessment of neuromotor function Quantitative measurements of tremor, eye–hand coordination (using an EKM), and, rapid alternating movements (tested using the DIADO system) were performed after the clinical evaluation. The results from the quantitative assessment of tremor have been reported elsewhere (Wastensson et al., 2006). All quantitative measurements were conducted by the same neuropsychologist (DL), who was blinded with respect to the subjects’ exposure status.
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
598
Fig. 1. The eurythmokinesimeter.
Fig. 2. The diadochokinesimeter.
2.4.1. Quantitative assessment of rapid pointing movements (eurythmokinesimetry) The EKM system was developed to quantify eye–hand coordination in a pointing task (Beuter et al., 1999a). This system is composed of two targets, one distal and one proximal, each divided into three electrically isolated concentric areas, and a pointer. Before the evaluation began, the subjects were asked to sit down in front of the system and hold the pointer like a pen. The rod with the mounted targets was inclined at a 308 angle, and adjusted to such a distance that the subject’s arm was only slightly bent when the pointer was in contact with the distal target. The distance between the targets, from center to center, was kept at a fixed distance of 25 cm. In the evaluation procedure, the subjects were asked to alternately touch the center of each target, as precisely and quickly as possible (Fig. 1). The subjects were familiarized with the equipment before testing. Each recording lasted 30 s and was repeated twice with both hands alternating, with a 15 s pause between each recording. Finally, the recordings were transformed to nine calculated measures (Table 3), used to characterize the performance (Beuter et al., 1999a). For the statistical analyses, four values (two trials and two targets) were obtained for each characteristic and averaged to a mean for each hand. Where a value for a trial or target was missing, the average of the remaining values was calculated.
2.4.2. Quantitative assessment of alternating movements of the forearms (diadochokinesimetry) The diadochokinesimeter was developed in the 1990s to quantify the performance of rapid alternating movements of the forearms (Beuter et al., 1994, 1999b). The subjects were asked to sit down in front of the system and firmly hold a soft sphere in each hand (Fig. 2). The spheres were fixed to flexible rods mounted on another rod and connected to optical encoders by bendable joints. Before the beginning of the evaluation, the rod was adjusted so that the subject’s elbows were flexed at a 908 angle, and were free from any obstacles. The subjects were asked to hold the spheres so that the palms of their hands were facing each other, and then execute alternating movements of the forearms as fast as possible, with both hands moving simultaneously. All subjects were given time to get familiarized with the equipment before the recordings begun. Each recording lasted 5 s and was repeated twice, with a 15 s pause between each trial. To characterize the performance, six measures were calculated (Table 4), as described elsewhere (Beuter et al., 1999b). For the statistical analyses, the mean for each hand over the two trials for each characteristic was used. Where the value for one trial was missing, the remaining value was used in the calculations.
Table 3 Definitions of measures used to characterize rapid pointing movements, recorded with a eurythmokinesimeter Characteristicsa
Definitions
Speed
The number of times the target was struck (the number of events on target), divided by the sum of the times taken to reach the target before each event (in events per second). Larger scores indicate faster performance. The proportion of events involving a strike on target area A. Larger scores indicate a more precise performance. The proportion of strikes involving a strike on target area B, C, or D. Larger scores indicate a more imprecise performance. The average number of contacts per event. Smaller scores indicate lower disposition to sideslip across target areas or multiple contacts in one target area, and, therefore, better performance. The number of contacts, less the number of target areas contacted (averaged over events). Measurement of the number of extra contacts after the initial contact when there are multiple contacts on a target area. Smaller scores indicate less tremor interfering with the performance. The average duration of transportation of the hand from one target to another. Smaller scores indicate better performance. The average total duration of contacts on the target. Smaller scores indicate shorter contact and, therefore, better performance. The constant, k, is calculated as the average over events of k = t/log (2A/W), where t = the transit time to the target, A = the distance between the two target centers = 25 cm, and W = the approximate distance between the location of the contact(s) and the target center (for events with contacts on only one target area, the midpoint of the minimum and maximum distances of points in the area from the center was used; for events with contacts in two adjacent target areas, the distance of the separator between the two areas was used). This constant k should be a measure of inherent ability, independent of the subject’s choice in the speed/accuracy tradeoff. Smaller scores indicate better performance. The standard deviation of intervals between events. Events from both targets and trials were used together to calculate a single standard deviation. Smaller scores indicate more regular and, consequently, better performance.
Precision Imprecision Unsureness Tremor
Transit duration Contact duration Fitts’ law constant
Irregularity a
Definitions of characteristics from Beuter et al. (1999a).
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
599
Table 4 Definitions of measures used to characterize rapid alternating movements of the forearms, recorded with a diadochokinesimeter Characteristicsa
Definitions
Duration Range
Mean duration of an oscillation (in seconds). Larger values are considered worse in fast cadence conditions. Mean of total angular displacement (pronation and supination) per cycle (in degrees). Larger values indicate greater range, and, therefore, better performance. Mean velocity for each cycle, averaged over all cycles (in degrees per second). Larger values are better in fast cadence conditions. From the second differences of the signal, the sum of all absolute values of negative accelerations, where the velocity generally increases from zero to its maximum value, and all positive accelerations, where the velocity generally decreases from its maximum value to zero. The larger the number the more irregular the performance. The mean absolute value of velocity relative to peak velocity for ascending and descending segments, and averaged across all cycles. The larger the number the less difficulty there is at the turn. Maximum slope of a regression line, fitted to seven successive data points and averaged over each ascending and descending part of all the cycles (in degrees per sampling point; multiplied by 200 to obtain degrees per second). Larger values are better in fast cadence conditions.
Velocity Smoothness
Sharpness Maximum slope
a
Definitions of characteristics from Beuter et al. (1999b).
2.5. Statistics
3.2. Eurythmokinesimetry
For group comparisons, Student’s t-test or Wilcoxon’s rank sum test (n < 20, or skewed data) was used. Spearman’s correlation coefficients were used to evaluate associations between the outcome variables from the EKM and DIADO systems, mercury exposure, and potential confounders (age, shift work, smoking). Associations between the quantitative measurements of neuromotor function and mercury exposure were examined using multiple linear regression analysis, which allows for adjustment for potential confounders. The exposure indices (U-Hg, U-Hgm5, U-Hgcum) were included separately in the model. Variability between trials within individuals was assessed by the coefficient of variation (CV = relative standard deviation). p-Values of <0.05 (for twotailed tests) are reported as statistically significant. Statistics were calculated with the SAS statistical package, version 8.2 (SAS, 1999).
3.2.1. Associations with mercury exposure There were no significant group differences between the exposed subjects and referents in any of the characteristics used to describe eye–hand coordination, as measured by the EKM system (Table 6). In the multivariate analyses, unsureness was significantly associated with U-HgC and U-Hgm5 among exposed subjects, with age, work schedule, and smoking included in the model. In the dominant hand, these associations were, however, due to a single outlier. In the nondominant hand, these findings were supported by a similar significant association in the entire group, with age, work schedule and smoking included in the model. Tremor in both hands was significantly associated with UHgC and U-Hgm5 in the multivariate analyses, but in the dominant hand, all associations were due to the same outlier mentioned above. In the nondominant hand, no association between tremor and U-HgC was found in the entire group, and in fact, the exposed workers had less tremor compared with the referents (Table 6). After taking outliers into account, no significant associations were found between previous Hg exposure (U-Hgcum) and any of the outcome variables from the EKM test.
3. Results 3.1. Clinical examinations In 16 subjects, 12 exposed subjects and four referents, abnormalities, albeit mainly minor deviations, were found in the neurological tests. Of these, 10 subjects were classified as having abnormal tremor (rest, postural, or intention tremor). The results from the clinical examinations including excluded persons are summarized in Table 5.
3.2.2. Associations between measures, trials and targets Correlations between measures in the entire study group (n = 64, one subject missing) were calculated. As expected, we found that speed was highly negatively correlated to transit duration (dominant hand, rs = 0.84; nondominant hand, rs = 0.89) and precision (rs = 0.77; 0.80). Speed was also highly negatively correlated to Fitts’ law constant (dominant hand,
Table 5 Results from clinical examinations of 58 exposed subjects and 35 referents Final group (n = 65)
2
Grip strength, dominant hand, kp/cm (range) Grip strength, nondominant hand, kp/cm2 (range) Rest tremor, % (n) Postural tremor, % (n) Intention tremor (finger–nose), % (n) Abnormal knee–heel test, % (n) Dysdiadochokinesia, % (n) Abnormal Romberg’s test, % (n) Gait disturbance, % (n) Hyporeflexia, % (n) Hyperreflexia, % (n) Reduced sensation, % (n) Reduced sense of vibration, % (n)
Excluded persons (n = 28)
Exposed subjects (n = 43)
Referents (n = 22)
Exposed subjects (n = 15)
Referents (n = 13)
1.23 (0.75–1.60) 1.19 (0.80–1.60) 2% (1) 12% (5) 12% (5) (0) (0) (0) (0) 9% (4) (0) 5% (4) 5% (4)
1.18 (0.85–1.50) 1.13 (0.80–1.45) (0) 14% (3) 9% (2) (0) (0) (0) (0) (0) (0) 5% (1) 5% (1)
1.23 (1.00–1.70) 1.13 (0.70–1.50) (0) 27% (4) 33% (5) 7% (1) (0) (0) (0) 13% (2) (0) 13% (2) 27% (4)
1.23 (1.05–1.60) 1.15 (0.90–1.50) (0) 31% (4) 46% (6) (0) (0) 8% (1) (0) 8% (1) (0) 23% (3) 23% (3)
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
600
Table 6 Results of measurement of rapid pointing movements, using the eurythmokinesimeter§ Characteristicsa
p-Valueb
Group comparisons Exposed (n = 42)c
subjects
Correlation coefficients (Spearman’s)
Referents (n = 22)
All (n = 64)
Exposed subjects (n = 42)
U-HgC
U-HgC
S.D.
Mean
S.D.
Dominant hand Speed (mm/s) Precision Imprecision Unsureness Tremor Transit duration Contact duration Fitts’ law constant Irregularity
1.10 0.48 0.69 1.38 0.22 0.77 0.24 0.14 0.15
0.33 0.20 0.17 0.34 0.21 0.19 0.20 0.03 0.10
1.06 0.55 0.68 1.46 0.24 0.78 0.23 0.14 0.13
0.25 0.23 0.20 0.27 0.15 0.19 0.16 0.03 0.07
0.62 0.19 0.74 0.34 0.78 0.85 0.87 0.85 0.51
0.06 0.02 0.07 0.07 0.03 0.01 0.04 0.02 0.07
0.06 0.18 0.11 0.10 0.15 0.02 0.09 0.01 0.04
0.15 0.15 0.16 0.15 0.10 0.22 0.01 0.23 0.08
0.05 0.03 0.07 0.04 0.02 0.05 0.01 0.07 0.23
Nondominant hand Speed (mm/s) Precision Imprecision Unsureness Tremor Transit duration Contact duration Fitts’ law constant Irregularity
1.05 0.47 0.78 1.52 0.24 0.81 0.24 0.15 0.14
0.34 0.19 0.13 0.30 0.15 0.20 0.19 0.03 0.07
1.00 0.51 0.74 1.56 0.32 0.82 0.26 0.15 0.14
0.27 0.25 0.17 0.40 0.21 0.23 0.19 0.03 0.07
0.62 0.42 0.35 0.64 0.13 0.89 0.69 0.92 0.97
0.04 0.01 0.12 0.09 0.09 0.03 0.00 0.05 0.05
0.03 0.03 0.17 0.09 0.06 0.04 0.04 0.03 0.05
0.16 0.08 0.26 0.03 0.01 0.17 0.07 0.18 0.02
0.01 0.00 0.15 0.22 0.12 0.04 0.03 0.05 0.12
§
U-Hgcum
U-Hgm5d
Mean
For an explanation of the characteristics, see Table 3. a Mean of four recordings. b t-test, adjusted for unequal variance. c One subject missing. d n = 39.
rs = 0.81; nondominant hand, rs = 0.86), a measure which is independent of the subject’s choice of speed/accuracy strategy. Furthermore, strong associations were found between unsureness and tremor (dominant hand, rs = 0.85; nondominant hand, rs = 0.76). Intraindividual variability for the four values (two trials and two targets) for each hand was calculated for all characteristics in the entire group (n = 64). It was low for most characteristics, with the CV ranging from 9% to 31%. However, the variability in tremor was high in both hands (dominant hand, 70%; nondominant hand, 83%), and for irregularity, the CV was 58% in the nondominant hand and >100% in the dominant hand. 3.2.3. Associations with age, shift work, and smoking habits In the entire study group (n = 64, one subject missing), increasing age was associated with a decrease in speed (dominant hand, rs = 0.33, p = 0.007; nondominant hand rs = 0.30, p = 0.02), and with longer contact duration between the stylus and metal targets (dominant hand, rs = 0.42, p = 0.0005; nondominant hand, rs = 0.44, p = 0.0003). In the nondominant hand, older subjects had more tremor (rs = 0.29, p = 0.02), but in contrast, a more precise performance (rs = 0.29, p = 0.02). Smokers had significantly more tremor ( p = 0.048) and a tendency to sideslip across target areas ( p = 0.02) in the dominant hand compared with nonsmokers. No significant differences were found between shift workers and daytime workers in any of the outcome variables from the EKM test. 3.3. Diadochokinesimetry 3.3.1. Associations with mercury exposure There were no significant overall group differences between exposed subjects and referents in any of the outcome variables from the test of rapid alternating movements of the forearms (Table 7). There were, however, some inverse associations
between Hg exposure, on the one hand, and duration, velocity, sharpness, and range, on the other (Table 7). In the multivariate analyses, an inverse association was found between velocity in the dominant hand and Hg exposure (U-HgC and U-Hgm5) among exposed subjects, and it remained statistically significant with age and work schedule included in the model. Furthermore, the ability to make fast turns (sharpness) was inversely associated with U-HgC and U-Hgm5 among exposed subjects in the dominant hand, and was close to significant with age and shift work included in the model. A further analysis comparing a subgroup of the exposed workers with the highest U-HgC (n = 22, U-HgC 5.9; median level) or subjects with the highest U-Hgm5 (n = 20, U-Hgm5 6.75; median level) with the referents, did not reveal any differences with regard to velocity or sharpness. Finally, range in the dominant hand was inversely associated with U-HgC and U-Hgm5, but these associations disappeared when a single outlier was removed. No significant associations were found between previous Hg exposure and the outcome variables from the DIADO test in the multivariate analyses. 3.3.2. Associations between measures and trials In the entire study group (n = 65), duration was highly negatively correlated to velocity (dominant hand, rs = 0.73; nondominant hand, rs = 0.70) and sharpness (rs = 0.76; 0.79). Moderate to strong associations were found between maximum slope and range (dominant hand, rs = 0.54; nondominant hand, rs = 0.60), and velocity (rs = 0.67; 0.68). For most characteristics, correlations between the first and second trials were high in both hands. Expressed as the CV, the intraindividual variation for duration, range, velocity, sharpness, and maximum slope was about 8% (range 6–12%). The variability in smoothness was, however, very high, and even when an outlier was removed the CV was >100% in both hands.
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
601
Table 7 Results of measurement of rapid alternating movements, using a diadochokinesimeter§ Characteristicsa
Exposed (n = 43)
Dominant hand Duration Range Velocity Smoothness (10 Sharpness Maximum slope
2
)
Nondominant hand Duration Range Velocity Smoothness (10 2) Sharpness Maximum slope §
p-Valuec
Group comparisons subjects
Correlation coefficientsb (Spearman’s)
Referents (n = 22)
All (n = 65)
Exposed subjects (n = 43)
U-HgC
U-HgC
U-Hgcum
U-Hgm5d
Mean
S.D.
Mean
S.D.
4.76 333.76 75.93 2.81 0.49 7.61
1.45 54.34 20.06 0.13 0.08 1.42
5.22 347.77 74.29 0.75 0.49 7.50
2.10 63.99 22.26 0.54 0.10 1.70
0.37 0.36 0.77 0.30 0.98 0.77
0.05 0.25* 0.19 0.01 0.19 0.02
0.23 0.24 0.40* 0.06 0.29 0.12
0.35* 0.19 0.24 0.03 0.21 0.11
0.26 0.11 0.29 0.00 0.26 0.06
4.78 333.06 74.85 0.58 0.50 7.31
1.44 55.99 18.38 0.45 0.08 1.36
5.20 335.03 71.22 0.77 0.50 7.05
1.99 54.79 19.58 0.93 0.09 1.59
0.33 0.89 0.46 0.37 0.77 0.50
0.07 0.04 0.08 0.15 0.09 0.07
0.25 0.03 0.20 0.01 0.19 0.00
0.37* 0.22 0.23 0.17 0.32* 0.00
0.26 0.03 0.08 0.19 0.18 0.08
For an explanation of the characteristics, see Table 4. a Mean of two recordings. b *p-value < 0.05. c t-test, adjusted for unequal variance. d n = 40.
3.3.3. Associations with age, shift work, and smoking habits A high age was associated with reduced sharpness (rs = 0.32, p = 0.008) in the nondominant hand (n = 65), but the association did not reach statistical significance in the dominant hand. Shift workers performed alternating movements significantly faster in both hands compared with daytime workers. Nonsmokers had a significantly greater range and, consequently, better performance than smokers in both hands. 3.4. Comparison between the clinical assessment and the quantitative tests The original group (n = 93) was used for comparing clinical findings with the outcome variables in the quantitative tests. First, all subjects with neurological signs (n = 31) were compared with the remaining subjects (n = 62). Subjects with neurological abnormalities were significantly older (mean age 47 years, compared with 39 years among the other subjects) but did not differ in any of the measurements in the EKM or DIADO tests. Interestingly, subjects with abnormal tremor (rest, postural or intention tremors) (n = 21) had significantly shorter duration, higher velocity, and maximum slope (better performance) in both hands than other subjects in the DIADO test. The difference remained significant for velocity and maximum slope when subjects with only postural (n = 15) or intention tremor (n = 18) were compared with the other subjects. No significant differences were shown for the EKM test. 4. Discussion 4.1. Discussion of findings Overall, this study does not indicate any significant adverse effects of long-term low level Hg exposure on certain measures of neuromotor function. The clinical examinations did not reveal more neurological abnormalities among exposed subjects than the referent group. Some findings in the EKM test, however, indicate that current mercury exposure may have an effect on eye–hand coordination (unsureness, tremor) in the nondominant hand. In
the DIADO test, no significant associations with Hg exposure were found, except for a possible effect on velocity in the dominant hand among exposed subjects only. A tendency towards more neurological abnormalities has been reported in studies of workers with high current Hg exposure, down to U-Hg levels of around 70 mg/gC (Miller et al., 1975; Ehrenberg et al., 1991; Urban et al., 1996). The abnormal findings reported include eyelid fasciculations (Miller et al., 1975), static tremor (Ehrenberg et al., 1991; Urban et al., 1996), hyperactive tendon reflexes (Miller et al., 1975), absence of ankle jerk (Urban et al., 1996), dysdiadochokinesis, difficulties with heel–toe gait, and rest and intention tremors (Ehrenberg et al., 1991). In accordance with the present study, no increase in neurological abnormalities has been shown among exposed workers at lower (U-Hg 20–25 mg/gC or mg/L) exposure levels (Chapman et al., 1990; Langworth et al., 1992). Effects on motor speed and coordination have been reported among workers with high or moderate Hg exposure (>U-Hg 25 mg/ gC or mg/L), with several studies reporting reduced performance on finger tapping (Miller et al., 1975; Langolf et al., 1978; Gu¨nther et al., 1996) and eye–hand coordination (Miller et al., 1975; Langolf et al., 1978; Roels et al., 1982, 1989; Williamson et al., 1982; Piikivi et al., 1984; Gu¨nther et al., 1996). The results from studies with lower exposure levels (U-Hg 25 mg/g or mg/L) are somewhat contradictory. In a study of 89 chloralkali workers with an average U-Hg of around 25 mg/gC (Langworth et al., 1992), no differences in finger tapping or eye– hand coordination were found between exposed subjects and referents. The performance in these tests was, however, negatively correlated to the relative number of U-Hg peaks >30 mg/L among exposed workers. In another study of chloralkali workers, with a mean U-Hg of 18 mg/g, no effect on eye–hand coordination or finger tapping was found, and surprisingly, the exposed workers performed better (Piikivi and Ha¨nninen, 1989). By contrast, reduced performance in finger tapping was shown among exposed subjects compared with the referent group in a study of 88 lamp factory workers with similar exposure levels as in the studies above (Liang et al., 1993).
602
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
A possible explanation for the discrepancy in observed effects among currently exposed workers may be different historical Hg exposure in the study populations. Several studies show an excess of abnormal clinical findings among previously exposed workers even long after cessation of exposure (Albers et al., 1988; Andersen et al., 1993; Powell, 2000; Letz et al., 2000; Frumkin et al., 2001). In addition, effects on finger tapping (Frumkin et al., 2001; Kishi et al., 1994), eye–hand coordination and manual dexterity (Frumkin et al., 2001; Kishi et al., 1994; Mathiesen et al., 1999; Powell, 2000) were shown. In these studies, however, the historical exposure levels were generally high, and a recent study on previously exposed workers with similar cumulative Hg exposure as for the workers in the present study showed no effect on motor speed and coordination, which is in accordance with our findings (BastPettersen et al., 2005). An association between reduced motor speed, as evaluated with finger tapping, and low exposure levels (U-Hg 10 mg/g) has been shown in a study of industrial workers including workers from the chloralkali industry (Lucchini et al., 2002) and in two studies on dental personnel (Ngim et al., 1992; Echeverria et al., 1998). These findings are, however, in contrast to the absence of such associations in another study of chloralkali workers with low exposure levels (Ellingsen et al., 2001). Our findings of possible effects on eye–hand coordination (evaluated using the EKM test) have little support from negative studies using other tests of manual dexterity, such as the grooved pegboard test (Ngim et al., 1992; Ellingsen et al., 2001) and the one-hole pins test (Echeverria et al., 1998). Impaired performance of rapid alternating movements has not been described at lower exposure levels, and as far as we know, quantitative measurement of these movements has not previously been applied in subjects exposed to Hg0. Consequently, our finding of a possible effect on velocity in the dominant hand must be confirmed by other studies. As expected, older subjects were slower, mainly due to longer contact with the targets, and had more tremor (nondominant hand) in the hand–eye coordination test (using the EKM system). We could not, however, confirm an association between age and more irregularity or high values on Fitts’ law constant, indicating a general impairment of performance by age, as described in another study (Beuter et al., 1999). Except for the reduced ability, among older subjects, to rapidly stop and start alternating movements (sharpness) in the nondominant hand, age was not significantly associated with any of the other outcome variables in the DIADO test. Possible explanations for the modest effect of age on these quantitative tests in contrast to other studies (Beuter et al., 1999) may be that our subjects were a selected group (i.e., they were healthier and all still of working age), or that the sample size was small. Shift work was not found to impair the ability to execute tasks demanding good eye–hand coordination or the performance of rapid alternating movements, which could have been expected. Our finding that shift workers were faster than daytime workers in the DIADO test could possibly be a ‘‘healthy workers effect’’ (i.e., shift workers may tend to move to daytime work when they get health problems). The fact that the shift workers had fewer abnormal findings in the clinical examinations (21% vs. 32% among daytime workers) speaks in favor of this explanation. In the EKM test, smokers had more tremor and tendency to sideslip across target areas compared with nonsmokers. In accordance with our findings, increased tremor related to smoking habits has been reported in other occupational studies using quantitative tests of hand steadiness and tremor (Ellingsen et al., 2001, 2006; Bast-Pettersen et al., 2004), and has been shown to affect a variety of hand activities (Louis, 2007). In a study of 297 subjects from the general population in Quebec, Canada, with low
level exposure to manganese (Beuter et al., 1999), no associations between smoking habits and any of the outcome variables from the DIADO test were shown. Consequently, our finding of a possible effect of smoking habits on range in the DIADO test has to be confirmed by other studies. 4.2. Aspects of validity The exposed subjects and the referents were similar in important background characteristics, such as age, smoking habits, and work schedule, indicating good comparability between groups. Other sources that may give a small contribution to Hg0 exposure, i.e., amalgam fillings or high fish consumption were similar among exposed chloralkali workers and referents. All workers were employed at the same plants, and the participation rate was high, indicating no major selection bias. We chose to exclude all subjects with diseases or medication that may affect neuromotor function, in order to reduce ‘‘noise’’ and to be able to detect more subtle effects that might be caused by Hg exposure. On the other hand, when the number of subjects was reduced, only few subjects with high U-Hg levels were left among our exposed subjects. Consequently, if there is a threshold, and if Hg exposure affects only subjects with the highest U-Hg levels, such an effect could not be detected. In fact, the ‘outlier’ mentioned above, with abnormal results in some EKM measures (unsureness, tremor), had experienced relatively high exposure in the last 5 years (mean 37 mg/gC). Since no other plausible explanation (family history of tremor, disease or medication) was found, Hg exposure may be considered as a possible cause to the abnormal results in this subject. In the present study, highly sensitive methods were used to evaluate possible effects of low level mercury exposure on neuromotor function. To be applicable in the occupational field, these methods must be reliable and simple to use, however. For the EKM test, most characteristics were stable over trials; however, two trials may be insufficient for tremor and irregularity, as these characteristics show high intraindividual variation. In the DIADO test, the recordings were made in fast cadence, with both hands moving simultaneously. This condition is anticipated to exacerbate any subclinical adverse effects on the CNS more than other conditions (Beuter et al., 1994); also, the characteristics used in our study have been shown to be consistent between trials in fast cadence (Beuter et al., 1999b). However, our results indicate that the number of trials (or duration of test) is insufficient for smoothness, which has high intraindividual variability. Our study group (n = 43 + 22) was small, but not extremely small. We performed power calculations on several of the outcome variables and found that the statistical power was moderate to good. For example, in the present study, the power was 80% (a = 0.05) to detect a 10% difference in mean Fitts’ law constant (EKM test) and sharpness (DIADO test), or a 20% difference in mean precision (EKM test) between the groups. The EKM system is anticipated to be similar to the finger–nose test. Surprisingly, the comparison between the clinical evaluation (intention tremor) and the quantitative measures from the EKM test showed no associations but is in accordance with another study (Beuter et al., 1999a). In the finger–nose test, the subject is asked to alternately touch the end of the nose with the tip of the index finger; and the examiner classifies the performance as normal or abnormal (slight or obvious tremor). As has been indicated by Notermans et al. (1994), these clinical tests are usually not sufficiently sensitive to detect subclinical changes in neuromotor function, and contain a considerable degree of subjectivity when criteria are applied (McKeown-Eyssen et al., 1990). In addition, marked interobserver variation has been
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
reported in the assessment of signs of neurological abnormalities (McKeown-Eyssen et al., 1990). In the present study, two physicians conducted the clinical examinations, and consequently, variability between observers may have affected the results. Moreover, the quantitative test measures several distinct aspects of the performance, and is, in contrast to the clinical evaluation, a continuous variable. There was also a certain time delay (about 15– 30 min) between the two tests. Finally, the two tests may not be completely comparable. In the finger–nose test, the subjects performed the test with their eyes closed, while the EKM test demands good function of the visual pathways. In the present study, all subjects were classified as normal in the clinical assessment of diadochokinesis; therefore, a comparison with the quantitative method (DIADO test) was not possible. We have no obvious explanation for our findings that subjects with tremor at clinical examination performed better in the DIADO test compared with subjects without clinical tremor. In conclusion, the present study reveals no significant adverse effects of exposure to Hg0 on certain aspects (eye–hand coordination, diadochokinesis) of neuromotor function at low exposure levels. The lack of agreement between clinical tests (finger–nose test) and quantitative assessment (EKM test) was surprising. In general, the quantitative systems used in this study were reliable and easy to handle. They should be suitable for studies of groups exposed to neurotoxic chemicals. However, certain modifications (duration of test, number of trials) may be useful to compensate for the high intraindividual variability in some outcome measures. Acknowledgements The authors would like to thank the participants in this study and the staff at the local Occupational Health service units for their assistance. We are grateful to Annika Claesson, for helpful assistance with collecting and handling the urine samples, and the late Andrejs Schu¨tz, head of the laboratory at the Department of Occupational and Environmental Medicine in Lund. Finally, we wish to thank Eva Andersson for statistical advice.
References Albers JW, Kallenbach LR, Fine LJ, Langolf GD, Wolfe RA, Donofrio PD, et al. Neurological abnormalities associated with remote occupational elemental mercury exposure. Ann Neurol 1988;24:651–9. Andersen A, Ellingsen DG, Mørland T, Kjuus H. A neurological and neurophysiological study of chloralkali workers previously exposed to mercury vapour. Acta Neurol Scand 1993;88:427–33. Bast-Pettersen R, Ellingsen DG, Hetland SM, Thomassen Y. Neuropsychological function in manganese alloy plant workers. Int Arch Occup Environ Health 2004;77:277–87. Bast-Pettersen R, Ellingsen DG, Efskind J, Jordskogen R, Thomassen Y. A neurobehaviorial study of chloralkali workers after the cessation of exposure to mercury vapor. Neurotoxicology 2005;26:427–37. Beuter A, Mergler D, De Geoffroy A, Carrie`re L, Be´langer S, Varghese L, et al. Diadochokinesimetry: a study of patients with Parkinson’s disease and manganese exposed workers. Neurotoxicology 1994;15:655–64. Beuter A, Edwards R, De Geoffroy A, Mergler D, Hudnell K. Quantification of neuromotor function for detection of the effects of manganese. Neurotoxicology 1999;20:355–66. Beuter A, De Geoffroy A, Edwards R. Quantitative analysis of rapid pointing movements in Cree subjects exposed to methylmercury and in subjects with neurological deficits. Environ Res 1999a;80:50–63. Beuter A, De Geoffroy A, Edwards R. Analysis of rapid alternating movements in Cree subjects exposed to methylmercury and in subjects with neurological deficits. Environ Res 1999b;80:64–79.
603
Chapman LJ, Sauter SL, Henning RA, Dodson VN, Reddan WG, Matthews CG. Differences in frequency of finger tremor in otherwise asymptomatic mercury workers. Br J Ind Med 1990;47:838–43. Clarkson TW, Magos L. The toxicology of mercury and its chemical compounds. Crit Rew Toxicol 2006;36:609–62. Echeverria D, Aposhian HV, Woods JS, Heyer NJ, Aposhian MM, Bittner AC Jr, et al. Neurobehavioral effects from exposure to dental amalgam Hg0: new distinctions between recent exposure and Hg body burden. FASEB J 1998;12:971–80. Ehrenberg RL, Vogt RL, Smith AB, Brondum J, Brightwell WS, Hudson PJ, et al. Effects of elemental mercury exposure at a thermometer plant. Am J Ind Med 1991;19:495–507. Ellingsen DG, Bast-Pettersen R, Efskind J, Thomassen Y. Neuropsychological effects of low mercury vapor exposure in chloralkali workers. Neurotoxicology 2001;22:249–58. Ellingsen DG, Bast-Pettersen R, Efskind J, Gjølstad M, Olsen R, Thomassen Y, et al. Hand tremor related to smoking habits and the consumption of caffeine in male industrial workers. Neurotoxicology 2006;27:525–33. Fawer RF, De Ribaupierre Y, Guillemin MP, Berode M, Lob M. Measurement of hand tremor induced by industrial exposure to metallic mercury. Br J Ind Med 1983;40:204–8. Fimbel EJ, Domingo PP, Lamoureux D, Beuter A. Automatic detection of movement disorders using recordings of rapid alternating movements. J Neurosci Methods 2005;146:183–90. Frumkin H, Letz R, Williams PL, Gerr F, Pierce M, Sanders A, et al. Health effects of long-term mercury exposure among chloralkali plant workers. Am J Ind Med 2001;39:1–18. Gu¨nther W, Sietman B, Seeber A. Repeated neurobehavioral investigations in workers exposed to mercury in a chloralkali plant. Neurotoxicology 1996;17:605–14. Kishi R, Doi R, Fukuchi Y, Satoh H, Satoh T, Ono A, et al. Residual neurobehavioral effects associated with chronic exposure to mercury vapour. Occup Environ Med 1994;51:35–41. Langolf GD, Chaffin DB, Henderson R, Whittle HP. Evaluation of workers exposed to elemental mercury using quantitative tests of tremor and neuromuscular function. Am Ind Hyg Assoc J 1978;39:976–84. Langworth S, Almkvist O, So¨derman E, Wikstro¨m BO. Effects of occupational exposure to mercury vapour on the central nervous system. Br J Ind Med 1992;49:545–55. Letz R, Gerr F, Cragle D, Green RC, Watkins J, Fidler AT. Residual neurologic deficits 30 years after occupational exposure to elemental mercury. Neurotoxicology 2000;21:459–74. Liang YX, Sun RK, Sun Y, Chen ZQ, Li LH. Psychological effects of low exposure to mercury vapor: application of a computer-administered neurobehavioral evaluation system. Environ Res 1993;60:320–7. Louis ED. Kinetic tremor: differences between smokers and nonsmokers. Neurotoxicology 2007;28:569–75. Lucchini R, Cortesi I, Facco P, Benedetti L, Camerino D, Carta P, et al. Neurotoxic effect of exposure to low doses of mercury. Med Lav 2002;93:202–14. Mathiesen T, Ellingsen DG, Kjuus H. Neuropsychological effects associated with exposure to mercury vapor among former chloralkali workers. Scand J Work Environ Health 1999;25:342–50. McKeown-Eyssen GE, Ruedy J, Hogg S, Read Guernsey J, Woods I. Validity and reproducibility of a screening examination for neurological abnormality in persons exposed to methylmercury. J Clin Epidemiol 1990;43:489–98. Meyer-Baron M, Schaeper M, Van Thriel C, Seeber A. Neurobehavioral test results and exposure to inorganic mercury: in search of dose-response relations. Arch Toxicol 2004;78:207–11. Miller JM, Chaffin DB, Smith RG. Subclinical psychomotor and neuromuscular changes in workers exposed to inorganic mercury. Am Ind Hyg Assoc J 1975;36:725–33. Ngim CH, Foo SC, Boey KW, Jeyaratnam J. Chronic neurobehavioural effects of elemental mercury in dentists. Br J Ind Med 1992;49:782–90. Notermans NC, Van Dijk GW, Van der Graaf Y, Van Gijn J, Wokke JHJ. Measuring ataxia: quantification based on the standard neurological examination. J Neurol Neurosurg Psychiatry 1994;57:22–6. Piikivi L, Ha¨nninen H. Subjective symptoms and psychological performance of chlorine-alkali workers. Scand J Work Environ Health 1989;15:69–74.
604
G. Wastensson et al. / NeuroToxicology 29 (2008) 596–604
Piikivi L, Ha¨nninen H, Martelin T, Mantere P. Psychological performance and long-term exposure to mercury vapors. Scand J Work Environ Health 1984;10:35–41. Powell TJ. Chronic neurobehavioral effects of mercury poisoning on a group of Zulu chemical workers. Brain Inj 2000;14:797–814. Roels H, Lauwerys R, Buchet JP, Bernard A, Barthels A, Oversteyns M, et al. Comparison of renal function and psychomotor performance in workers exposed to elementary mercury. Int Arch Occup Environ Health 1982;50:77–93. Roels H, Abdeladim S, Braun M, Malchaire J, Lauwerys R. Detection of hand tremor in workers exposed to mercury vapor: a comparative study of three methods. Environ Res 1989;49:152–65. Rohling ML, Demakis GJ. A meta-analysis of the neuropsychological effects of occupational exposure to mercury. Clin Neuropsychol 2006;20:108–32. Sa¨llsten G, Barrega˚rd L, Ja¨rvholm B. Mercury in the Swedish chloralkali industry—an evaluation of the exposure and preventive measures over 40 years. Ann Occup Hyg 1990;34:205–14.
SAS Institute, Inc. Statistical software version 8. SAS Institute, Inc., Cary, NC, USA, 1999. Thorngren KG, Werner CO. Normal grip strength. Acta Orthop Scand 1979;50:255–9. Urban P, Lukas E, Benicky L, Moscovicova E. Neurological and electrophysiological examination of workers exposed to mercury vapors. Neurotoxicology 1996;17:191–6. Wastensson G, Lamoureux D, Sa¨llsten G, Beuter A, Barrega˚rd L. Quantitative tremor assessment in workers with current low exposure to mercury vapor. Neurotoxicol Teratol 2006;28:681–93. WHO (World Health Organization). Concise international chemical assessment document 50. Elemental mercury and inorganic mercury compounds: human health aspects. WHO, Geneva, Switzerland, 2003. Williamson AM, Teo RKC, Sanderson J. Occupational mercury exposure and its consequences for behaviour. Int Arch Environ Health 1982;50: 273–86.