Effects of Exposure to Elemental Mercury on the ...

Effects of Exposure to Elemental Mercury on the Nervous

Downloaded by [Reprints Desk Inc] at 04:49 21 December 2015

System and the Kidneys of Workers Producing Natural Gas

PETER J. BOOCAARD Shell International Chemicals Department of Molecular Toxicology Shell Research and Technology Centre Amsterdam, The Netherlands AUCUS-TINUS A. J. HOUTSMA Occupational Health Service Zuid-Drenthe Emmen, The Netherlands

H. LOUIS JOURN~E Biomedical Engineering and Medical Research Bedum, The Netherlands N l C O J. V A N SITTERT Shell International Chemicals Department of Molecular Toxicology Shell Research and Technology Centre Amsterdam, The Netherlands

JSCentre for

ABSTRACT. Early signs of alterations in renal and neurological functions were studied in three groups of workers who were exposed to different levels of mercury that were below the current biological exposure index of 35 pg/g creatinine. There were no differences among the three study groups with respect to either motor nerve conduction velocity or tremor frequency spectra of physiological tremors. Also, no significant correlations were found between the results of the neurological tests and any of the present or historical biological monitoring data. In contrast, N-acetyl-IJ-D-glucosaminidase was increased significantly in the group with the higher exposure, compared with either the lower-exposure or control groups. N-acetyl-B-D-glucosaminidasewas correlated strongly with mercury concentration in urine and was correlated weakly with historical biological monitoring data; however, there was no correlation with duration of exposure. These results suggest that after exposure to mercury at levels below the biological exposure index, a transient increase in N-acetyl-IJ-D-glucosaminidasecan be observed, but i s not an early indicator of developing renal dysfunction.

CERTAIN NATURAL GAS FIELDS contain substantial amounts of mercury. During the production of such fields, workers may be exposed to elemental mercury because it is produced with the gas. Occupational exposure to mercury vapors may lead to renal and neurological alteration^.'-^ Sensitive biochemical tests have been used to detect early signs of renal alterations that have occurred after exposures that caused urinary mercury concentrations to exceed 50 pg/g ~ r e a t i n i n e .Abnormal ~,~ tremors and other neurological effects may even occur after exposures that result in urinary mercury concentrations as low as 35 pg/g ~ r e a t i n i n e .The ~ - ~ American Conference of Governmental Industrial Hygienists recently adopted 108

a biological exposure index (BEl) for urinary mercury concentrations of 35 p.Lg/gcreatinine in preshift sample^.^ In this study, motor nerve conduction velocity (MNCV) and frequency spectra of physiological tremors were investigated as tests of peripheral neurotoxicity in three groups of workers who experienced different levels of exposure to elemental mercury; controls were also included. Biochemical parameters for renal function that may be indicative of early alterations of the kidney were also studied in these subjects. The first objective was to investigate whether the neurological or biochemical tests revealed any differences among the groups of workers. The second aim sought to determine whether any of the neurological or renal parameters Archives of Environmental Health

correlated with the occupational history of exposure to mercury.


Method and Material Population and study design. The study was conducted at a large gas-production site in the northeastern part of The Netherlands. Gas-production and treatment installations are halted for inspection and maintenance operations annually. Normally this shutdown occurs in summer, when demand for gas is low. The subjects of the high-exposure group (HIGH) were selected from the employees of an industrial cleaning company. A survey of the personnel files and interviews were used to select 18 men who had been involved regularly in cleaning activities at the gas-production sites. Two groups of employees with relatively low exposures were identified. The first group (LOW-1) comprised 1 laboratory technician and 8 men charged with the supervision of maintenance operations. These workers had a potential chronic exposure to low concentrations of mercury. Incidentally, they may have been also exposed to high concentrations of elemental mercury; in these cases, however, respiratory protection was usually applied. Their occupational histories did not indicate potential exposure to other heavy metals or other nephrotoxicants. The second group of employees with relatively low exposure to mercury (LOW-2) comprised 13 men who dismantled and reassembled the installations. These workers were also exposed incidentally to low concentrations of mercury. A control group (CONTROL) comprised 19 employees, of whom 13 were charged with inspections and minor technical maintenance operations at the gas-production site under study. The remaining 6 workers were employed at a plant that produced and treated gas from fields at more western locations where, up to now, no mercury was found. Persons who had a known history of nonoccupational neuropathies or who had a history of disorders with potential renal sequelae (e.g., urinary-tract disease, diabetes, hypertension [defined as untreated diastolic blood pressure over 95 mm Hg]) were excluded from the study. A self-administered questionnaire that contained questions about occupational and medical history (including dentistry), demographic characteristics, and lifestyle, was completed by each participant. Prior to testing, all questionnaires were reviewed for completeness and accuracy by the medical officer of the plant. Information gleaned from the questionnaires did not lead to exclusion of workers from the study. All workers in LOW-1, LOW-2, and CONTROL groups worked within the same company. The occupational history records indicated that the operators had experienced no job transfers during their employment with the company that would have placed them in another study group. Effects on the renal tubules were investigated by determinationof urinary N-acetyl-R-D-glucosaminidase (NAG) and B,-microglobulin (R2M).Urinary albumin and total protein determinationswere used to study the MarchIApril 1996 Wol. 51 (No.2)]

integrity of the renal glomeruli. Effects on the nervous system were determined by assessment of peripheral nerve conduction and tremor frequencies of the arm. All participants were asked to refrain from drinking coffee for at least 2 h before of the neurological tests were conducted. On the day the neurological tests were performed, urine and blood samples were also collected. Blood was collected by puncture from the antecubital vein in two 10-ml mercury-free vacuum tubes. The contents of one tube were used on the same day to determine glucose concentration; the remaining tube was stored frozen at -20 "C until the time of mercury analysis. Each participant was asked to ingest 3 g of sodium carbonate tablets 2 h prior to collection of the urine sample; this was done to produce a urinary pH that exceeded 5.5 because, in more acidic urine, R2Mis degraded in the bladder.1° Spot urine samples were collected in mercury-free polyethylene containers, and pH was determined immediately after voiding. Samples with a pH of less than 5.5 were not analyzed for R2M.To correct for intraindividual variations in urinary mercury excretion, we asked all participants to deliver, in addition to this urine sample, four spot urine samples collected during the mornings of the 4 work days that followed. The concentration of mercury in urine (HgU) was the average of the concentrations measured in these samples. Air measurements. Potential exposure to mercury was measured occasionally during the study by both stationary and personal air monitoring. For stationary air monitoring, we used a direct-readingapparatus (i.e., jer6me 41 1 mercury-vapor analyzer and Bacharach mercury sniffer MV2). Personal air monitoring was conducted with 3M mercury-vapor badges, which were analyzed for mercury content by 3M (Analytical Service, Occupational Health and Safety Products; St. Paul, Minnesota). Neurological tests. The MNCV was measured according to a standard procedure12 performed routinely at the University Hospital Groningen. The MNCV was obtained from the ulnar nerve-between the points olecranon and wrist-where the ulnar nerve was stimulated by application of a current pulse of 15 mA and 200-ms width. The M-wave response was obtained from electromyographic recordings at the hypothenar. Furthermore, nerve conduction times of the median nerve over the wrist were measured to identify isolated disorders (e.g., carpal tunnel syndrome). Limb temperature was checked to be at least 31 "C. Tremor recordings were completed bilaterally by accelerometers (weight = 18 g) fixed to the dorsum of the medial phalanx of the middle fingers, as well as by surface electromyograms (EMGs) of extensor and flexor groups in the forearm. The EMG signals were converted into tremor signals by subsequent steps of demodulation, analogous with digital conversion and digital processing, thus resulting in power spectra.13-15 The power spectra of the processed forearm EMGs were investigated under three conditions: (1) at rest, with the arms hanging down (i.e., rest tremor: REST-R and REST-L for the right and left arms, respectively); (2) with both arms stretched 109


forward horizontally (i.e., action tremor: ACT-R and ACT-L for the right and left arms, respectively); and (3) with bowed arms at the sustained end of the finger-nose test, with closed eyes (i.e., intentional tremor: INT-R and INT-L for the right and left arms, respectively). Blood and urine analysis. Concentrations of mercury in blood (HgB) and HgU were determined in duplicate by atomic absorption spectrometry (AAS) (Perkin Elmer, model 4001, using the cold-vapor technique with SnCl as reductant. Blood and urine samples were treated with potassium permanganate in sulfuric acid for 1 h to oxidize all mercury-containing compounds into mercuric ions. After excess permanganate was eliminated with hydroxyammonium chloride, the sample (Perkin Elmer MHS-20 mercury hydride system) was reduced and its mercury content measured. Commercial blood (Kontrollblut, Behringwerke AG and Seronorm Trace Elements [Nyegaard]) and urine (Lanonorm-Metalle, Behringwerke AG) samples with assigned values were used as calibration standards for mercury in blood and urine, respectively. The accuracy of the results, compared with reference materials with assigned values, exceeded 95%. The renal parameter albumin was determined by electroimmunoassay, and B,M, NAG, and TP were determined as reported previously."j Glucose in serum and creatinine in urine were determined on a Hitachi 705 auto analyzer, with standard clinical chemical techniques. Historical exposure. In addition to the data collected in the current study, historical biological monitoring data were available. Since 1969, exposure to elemental mercury had been monitored biologically by quarterly determination of HgU. A more systematic biological monitoring program was initiated in 1982. In this program, urine samples were collected from all employees with potential exposure to mercury at the beginning, during, and at the completion of inspection and maintenance operations-in addition to quarterly samples. Subsequent to 1982, urinary creatinine was determined routinely, and HgU was expressed as pg/g creatinine to correct for variations in urinary output. The workers in the HIGH group spent an average of 5.7 2.8 y (standard deviation [SO])in the biological monitoring program. The workers who were selected for the LOW-1 and LOW-2 groups participated in the biological monitoring program for 9.7 f 3.5 y (SD) and 10.3 f 3.9 y (SD),respectively. job titles of the workers in the control group (CONTROL) were not associated with exposure to mercury; therefore, these employees had not been engaged previously, on a regular basis, in the biological monitoring program for mercury. However, for several of them, one or more isolated urinary mercury results were available from measurements made occasionally when they had been involved in operations with supposed potential exposure to mercury vapors. The results from these occasional surveys did not differ from those obtained in the control populations for the same period.lb Historical urinalysis records for the production site were used to define the historical exposure indices (Table 1).

*

110

Table 1.-Historical

Mercury Exposure Indices

~~

Index

Definition

Cumulative urine mercury (CUM)

Measure of integrated exposure calculated as the sum of the quarterly urinary mercury concentrations for each person* (VgA).

Duration of exposure (DUN

The number of quarters in which a person had detectable amounts of mercury in the urine (quarters)t.

Maximal exposure (MAX) The highest recorded historical urinary mercury concentration

(pgA). Average of historical quarterly Mean urinary mercury concentrations (HHgU) urinary mercury concentrations for each person (@I)*. *If more than one reading was recorded for a person in a given quarter, the average of all his values was used for that quarter. tThe limit of detection was I 2 @I.

Calculations and statistics. Statistical analysis was performed with the aid of the mainframe version (version 6.07) of the SAS statistical software package, with a < .05 as the level of significance. Each parameter was tested for normality with the Kolmogorov-Smirnov test. The results of present HgU, INT-R, INT-L, and all biochemical parameters (except NAG) showed a Gaussian distribution after logarithmic transformation. The distributions of possible confounders in the study groups and the control group were compared, using Fisher's two-tailed exact test for discrete variables (i.e., fish, alcohol and coffee consumption, smoking habits, recent dental restorations, and medication use) and Wilcoxon's rank-sum test for continuous variables (i.e., age and underarm length). The influence of the possible confounders on the various biochemical and neurological variables were evaluated by F test, with stepwise regression analysis. The neurological and biochemical variables were evaluated by F test, with the stepwise regression analysis. The neurological and biochemical variables were modeled after logarithmic transformation if necessary, and the influence of the potential exposure to mercury was assessed by multivariate regression analysis. Multivariate regression analysis was also performed on the results for the biochemical and neurological variables of the entire population (i.e., all subjects of the study groups and control group), with the results for HgU and HgB as independent variables. Results Study populations. The LOW-1 and LOW-2 study groups were pooled into a single study group characterized by low potential exposure (LOW). This was done because the results for all metal, biochemical, and neurological determinations did not differ signifiArchives of Environmental Health

Table 2.4haracteristics of the Three Exposure Groups

High

Exposure group Low

Control

18 33.8 38.3 13 (72%) 8.2 15 (83%) 12.8 16 (89%) 0.8 16 (89%) 7.7 6 t (33%) 0.8 3 (1 7%)

22 42.9* 37.2 1 1 (500/,) 7.4 17 (77%) 7.4 19 (86%) 0.7 21 (96%) 4.8 9 (41%) 0.7 5 (31%)

19 35.7 37.6 1 1 (58%) 7.1 18 (95Yo) 12.1 18 (95%) 0.7 18 (95%) 5.1 10 (53%) 0.6 0 (0%)

Characteristic I

No. of subjects Mean age (y) Mean underarm length (cm) No. of smokers Smoking habits (cigarettesld) No. of alcohol users Alcohol consumption (glasseslwk) No. of fisheaters Fish consumption (meals/wk) No. of coffee users Coffee consumption (cups/d) No. of persons with recent dental restorations No. of amalgam fillings No. of medication users I


*Significantly different from high-exposure and control groups (Wilcoxon’s rank-sum test). tsignificantly different from control group (Fisher’s exact test).

Table 3.-Present

and Historical Mean Mercury Concentrations

Parameter

High

ExDosure erouD Low

Present mercury concentrationst 23.7* (17; 3.5-71.9)t 4.1* (5; 0.6-8.8) 3.55 (3; 0-9) 1.5 (0; 0-1 6)

Control

2.4 (2; 0.5-6.8) 2.2 (2; 0-8)

Historical mercury concentrationst CUM (pg/l) DUR (quarter) MAX (pg/l) HHgU (pg/l)

416; (298; 7-1 014) 8.8* (9: 1-1 5) 130. (120; 7-290) 41.8; (41; 7-72)

199$ (1 30; 28-855) 10.7* (10; 3-20) 57* (35; 12-193) 17.3* (12; 7-53)

9.1 (3; 1-58) 1.9 (1; 1-7) 4.9 (3; 1-20) 3.6 (3; 1 - 8 )

Notes: HgU = concentration of mercury in urine, HgB = concentration of mercury in blood, CUM = cumulative urine mercury, DUR = duration of exposure, MAX = maximal exposure, and HHgU = mean urinary mercury concentration (cf. Table 1). *Significantly different from low-exposure and control groups (MANOVA). tValues are means; numbers within parentheses are median and range, respectively. *Significantly different from control group (MANOVA). §Significantly different from low-exposure group (MANOVA).

cantly for LOW-1 and LOW-2 groups. In fact, the individuals in this LOW group might be considered as random samples from the same population (unpaired two-tailed t test). The characteristics of both study groups and the control group are shown in Table 2. The groups were matched well for all variables, except for age and number of persons with recent dental restoration. Whenever appropriate, the test results were analyzed after correction was made for age or dentistry. Environmental monitoring. During the current study, stationary air monitoring was conducted on 2 different days at 15 different locations on those occasions when exposure was anticipated (e.g., during opening and cleaning of tanks and filters and during installations). Air concentrations ranged from 10 to 1 500 pg Hg/m3 (median = 67 pg Hg/m3). At locations where no exMarch/April 1996 Wol. 51

(No.2)]

posure or only very little exposure to mercury was expected (e.g., control rooms, canteens), concentrations were not detectable or they were low (e.g., cabins of vacuum cleaner trucks); concentrations in these areas ranged from 0 to 6 pg Hg/m3 ( n = 8). For seven different operations during which potential exposure to mercury occurred (e.g., cleaning of tanks, scrubbers, and filters), personal air measurements were conducted on 10 different individuals. Nine of these individuals (all of whom were from the HIGH exposure group) used respiratory protection, and 1 (LOW group) did not use a respirator. The potential 8-h time-weighted-average (TWA) exposure ranged from 33 to 781 pg H u m 3 (median = 88 pg/m3). Biological monitoring. The results of the biological monitoring are shown in Table 3. The current blood mercury concentrations in the high-exposure workers 111


Median HgU (pg/L)

76

77

78

79

80

81

82

83

84

85

86

Year Fig. 1. Overview of historical biological monitoring data for the period between 1976 and 1986 for the high-exposure group (circles) and the low-exposure group (squares). Representedare the group median values, calculated from all HgU measurements in a specific quarter. Since 1982, individuals may have been monitored biologically more often than quarterly; for these cases, we used the individually averaged results to calculate the median.

values in either the low-exposure (i.e., 5 pg/g and 12 pg/l, respectively) or control group (i.e., 2 pg/g and 3 pg/l, respectively). The results of the present HgU and

2

. : :1

I

0

0 0

10

20

30

40

50

60

70

80

90

100

HGU (pg/g CREATININE) Fig. 2. Correlation of urinary NAG activity with current HgU in all workers (i.e., both study groups and control group). The dotted lines represent the tolerance limits of the 95% confidence interval. The Pearson correlation coefficient i s .53 (p < .0001).

were significantly higher than concentrations found in the low-exposure group. The present (HgU) and mean historical urinary mercury concentrations (HHgU) in the high-exposure group were 17 pg/g creatinine and 41 pg/l (median values), respectively. These concentrations were significantly higher than the corresponding 112

HHgU in the low-exposure workers were also significantly (albeit slightly) elevated, compared with controls. The historical biological monitoring data showed that the workers in the low-exposure group had past exposures to much higher levels of metallic mercury (Fig. 1). Nevertheless, cumulative exposure (CUM [see Table 1I) and maximum historical exposure were more than twice as high in the high-exposure group, compared with the low-exposure group. Sixty-three percent of the high-exposure and 14% of the low-exposure workers had exceeded a peak urinary mercury concentration of 100 pg/l at least once, whereas in 41% of high-exposure workers, the mean historical urinary mercury excretion exceeded 50 pg/l. Biochemical tests. The results of the biochemical tests are shown in Table 4 (top). Both I3 M and NAG were increased significantly in the hi&-exposure group, compared with the low-exposure group. It should be noted, however, that the individual results of the workers in both groups were within the tolerance limits of the 95% confidence interval, as determined in workers who were not exposed occupationally.16There were no statistically significant differences for any of the tests between the low-exposure and control groups. A good correlation existed between NAG and the present HgU values (Pearson correlation coefficient = .53, p < Archives of EnvironmentalHealth

0

’

0

ZOO

400

600

800

-

1000

1200


CUM (PdY 1) Fig. 3. Correlation of 6,M with cumulative urinary mercury (CUM) in all workers (i.e., both study groups and control group). The dotted lines represent the tolerance limits of the 95% confidence interval. The Pearson correlation coefficient is .33 $I = .015).

the exposure during the annual maintenance operations remains relatively high, and at some locations the median air concentrations even exceed the current Dutch exposure limit value of 50 pg/m3. Three of 18 industrial cleaners (i.e., high-exposure group) in this study had a present HgU above the BE1 of 35 pg/g creatinine. In other studies, HgU values above this BE1 were associated with neurological (i.e., decreased MNCV and increased tremors) and renal (i.e., increased NAG) e f f e ~ t s . ~ , ~ In ~ *our ” , ~study, ~ the average present HgU in the high-exposure group was 24 pg/g creatinine (range = 3.5-72 pg/g), and no significant differences between this group and either the low-exposure or control groups were found for any of the neurological parameters. In contrast to our findings, significant increases in tremor frequencies were found in 18 and 21 clinically asymptomatic chloralkali workers for whom mean HgUs of 23 pg/l and 36 pg/g creatinine, respectively, were determined.6*8 Neurological effects from longterm exposure to high concentrations of mercury vapor may be evident several years after exposure.” Ellingsen et al.2 recently assessed neurological parameters in 77 chloralkali workers who had been exposed to approximately 59 pg Hg/m3 for an avera e period of 7.9 y. Although the exposure had ceased or 12.3 y (average), significant correlations with neurological functions and historical biological monitoring parameters were found., In our study, however, there were no significant correlations between neurological parameters and present or historical biological monitoring. A possible explanation may be that the chloralkali workers in these studies were exposed continuously to mercury vapors, whereas the workers in our study were exposed intermittently during periods of several weeks once a year. In another study, subjective symptoms, psychological performance, and renal parameters (i.e., albumin and NAG) were investigated in 60 workers who had longterm low-mercury-vapor exposure and an average HgU of 18 pg/g creatinine (range = 3.4-56 pg/g).5*19 Although this study is in accordance with ours with respect to the lack of effects on neurological parameters, no difference between the exposed workers and controls was reported with respect to NAG9; however, we measured a significantly higher NAG activity in the high-exposure group, compared with either the lowexposure group or the controls. The workers in our , in high-exposure group also excreted slightly more RM their urine than did the low-exposure group, but there was no significant difference with respect to controls. No significant differences for any of 7 renal parameters, including NAG and B,M, were found between a group of 89 workers exposed to elemental mercury for an average of 13.5 y (mean HgU of 25.4 pg/g creatinine) and a group of 75 control workers who had no occupational exposure to m e r ~ u r y .Nonetheless, ~ a clear relationship between present HgU and NAG was found. This finding (i.e., no significant increase in NAG but a significant correlation between NAG and HgU) was confirmed in a recent study of 50 workers whose average exposure to mercury vapors was 11 y and whose mean HgU was 22 pg/g ~reatinine.~ Also, in our

!

.0001 [Fig. 21). No significant correlations were observed between present HgU and any of the other renal parameters. Significant (albeit much weaker) correlations were found between historical biological monitoring parameters (i.e., CUM and HHgU) and NAG and B,M. The strongest correlation was found with CUM (i.e., Pearson correlation coefficients of .31 [p = .024] for NAG and .33 [p = .0151 for B,M [Fig. 31). Neurologicaltests. To exclude tremors resulting from relative hypoglycemia, we measured serum glucose in all participants, and all individuals were normoglycemic. The results of MNCV and highest peak frequencies of the rest, action, and intention tremor (of both right and left arms) for the three exposure groups are shown in Table 4 (bottom). There were no statistically significant differences among the three study groups for any of the parameters tested. No statistically significant correlations were found between the results of the neurological tests and either the present or the historical biological monitoring parameters.

Discussion Since 1963, natural gas has been produced from fields in the northeastern part of The Netherlands; substantial amounts of mercury are present in gas from these fields. Workers who are employed at these locations are potentially exposed to elemental mercury vapors, especially during inspection and maintenance operations that require opening of the installations. Since 1969, workers with potential exposure to mercury have undergone regular biological monitoring. Hygienic measures and more stringent application of personal protective equipment have contributed to a decrease in mercury exposure over the years (Fig. 1). Nevertheless, as shown by the air measurements in the present study, March/Aprill996 Wol. 51 (No. 2)]

113

Table 4.-Results

of Biochemical and Neurological Tests ~

Exaosure erouD Parameter

High

Low

Control

Biochemical tests* Albumin (mg/g)

K%$

Total protein (mg/g)

7.8 (6.7;2.4-25.7) 102t (95;46-233) 4.52* (2.9;1.7-1 1.7) 40.6 (35;18-1 01 )

6.6 (4.8;2.4-20.2) 76 (64;26-248) 2.67 (2.4;0.4-6.2) 45.4 (32;11-1 70)

15.6 (9.2;1.5-68.5) 114 (83;32446) 3.27 (2.7;1 .O-11.2) 53.8 (45;17-207)

NeurologicaItests*


MNCV (m/s) REST-L (Hz) REST-R (Hz) ACT-L (Hz) ACT-R (Hz) INT-L (Hz) INT-R (Hz)

56.1 (57.5;45-73) 5.94 (4.9;2.9-10.4) 6.49 (6.5;1.4-1 0.3) 6.76 (6.8;2.3-12.0) 6.61 (97.05;2.2-10.3) 5.15 (4.9;4.2-8.0) 5.09 (4.85;3.9-8.0)

59.8 (60;49-70) 6.60(6.8;2.9-10.5) 6.45 (5.95;2.3-10.2) 7.75 (7.55;1.6-1 1.4) 7.39 (7.4;2.0-12.3) 5.50 (5.3;4.2-10.6) 5.38 (5.2;3.7-10.8)

55.8 (55;46-67) 6.64 (6.8;3.6-10.1) 6.34 (6.0;3.9-9.2) 7.00 (7.7;2.5-9.5) 6.67 (7.5;2.4-9.0) 5.39 (5.6;3.9-6.2) 5.22 (5.2;3.9-6.7)

Notes: P,M = beta-2-microglobulin; NAG = N-acetyl-P-D-glucosaminidase; MNCV = motor nerve conduction velocity; REST-L = at rest, with arms hanging down (i.e., rest tremor-for left arm); REST-R = at rest, with arms hanging down (i.e., rest tremor-for right arm); ACT-L = both arms stretched forward horizontally (i.e., action tremorfor left arm); ACT-R = both arms stretched forward horizontally (i.e., action tremor-for right arm); INT-L = bowed arms at the sustained end of the finger-nose test, with closed eyes (i.e., intentional tremor for left arm); and INT-R = bowed arms at the sustained end of the finger-nose test, with closed eyes (i.e., intentional tremor for right arm). *Mean (median; range). tsignificantly different from low-exposure group (MANOVA). *Significantly different from low-exposure and control groups (MANOVA).

study we found a highly significant correlation between present HgU and NAG (Fig. 2). Interestingly, the correlation between NAG and the historical biological monitoring parameters, CUM and HHgU, were much weaker, and there was absolutely no correlation between NAG and duration of exposure. This suggests that the increase in NAG may be a transient effect that is related to recent exposure, and that such increase may not necessarily be an early indicator of persistent renal alterations and even less of renal dysfunction. Again, the differences between our study and the studies in chloralkali workers may be explained by the intermittent exposure to mercury of our workers. The suggestion that recent exposure to mercury vapor may lead to a transient increase in NAG is supported by the study of 77 chloralkali workers referenced In this study, there was no difference in renal parameters between the exposed workers and controls; also, there was no correlation between renal parameters and any of the exposure-related indices. However! an increased prevalence of elevated renal test results was found among workers for whom exposure to mercury had ceased only shortly before the study, and there were indications that elevations of NAG, induced by exposure to mercury vapors, were reversible. In conclusion, no differences in MNCV or tremor frequencies were found among maintenance workers with relatively high, intermittent exposure to mercury vapors, operators with low mercury exposure, and control workers. As well, there was no significant correlation between these neurological parameters and either pre114

sent or historical biological monitoring results. With regard to renal parameters, increased urinary NAG activities were found in the maintenance workers, compared with either the operators with low exposure or the controls. There was a strong correlation between NAG and HgU, but NAG was correlated only weakly with historical biological monitoring data; a correlation was not found between NAG and duration of exposure. These results suggest that after exposure to mercury vapor, a transient increase in NAG can be observed, but such increase is not an early sign of developing renal dysfunction.20

* * * * * * * * * * Submitted for publication December 9, 1994;revised; accepted for publication April 6, 1995. Requests for reprints should be sent to P. 1. Boogaard, Ph.D., Shell International Chemicals, BV, Dept. Molecular Toxicology, SRTCA, P.O. Box 38000,1030 BN Amsterdam, The Netherlands.

* * * * * * * * * * References

1. Ehrenberg RL, Vogt RL, Smith AB, et al. Effects of elemental mercury exposure at a thermometer plant. Am J Ind Med 1991; 19:495-507. 2. Ellingsen DG, Merrland T, Andersen K, et al. Relation between exposure-related indices and neurological and neurophysiological effects in workers previously exposed to mercury vapour. Br j Ind Med 1993;50:73644. Archives of Environmental Health


3. Cdrdenas A, Roels H, Bernard AM, et al. Markers of early renal changes induced by industrial pollutants. 1. Application to workers exposed to mercury vapour. Br J Ind Med 1993; 50:17-27. 4. Langworth S, Elinder CG, Sundquist KC, et al. Renal and immunological effects of occupational exposure to inorganic mercury. Br J Ind Med 1992; 49:394401. 5. Piikivi L, Hanninen H. Subjective symptoms and psychological performance of chlorine-alkali workers. Sci J Work Environ Health 1989; 15:69-74. 6. Chapman LJ, Sauter SL, Henning RA, et al. Differences in frequency of finger tremor in otherwise asymptomatic mercury workers. Br J Ind Med 1990; 47S38-43. 7. Soleo L, Urbano ML, Petrera V, et al. Effects of low exposure to inorganic mercury on psychological performance. Br J Ind Med 1990; 47:105-09. 8. Verberk MM, S a l k HJA, Kemper CH. Tremor in workers with low exposure to metallic mercury. Am Ind Hyg Assoc J 1986; 47:559-62. 9. ACCIH. 1994-1995 Threshold limit values for chemical substances and physical agents and biological exposure indices. Cincinnati, OH: ACGIH, 1994. 10. Schardijn G, Statius van LW, Swaak AJC, et al. Urinary 8,microglobulin in upper and lower urinary tract infections. Lancet 1979; iS05-07. 11. Alberts JW, Kallenbach LR, Fine LJ, et al. Neurological abnormalities associated with remote occupational elemental mercury exposure. Ann Neurol 1988; 24:651-59.

March/April 1996 Wol. 51 (No.2)]

12. Kimura J. Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice. Philadelphia, PA: F.A. Davis Co. 13. lournee HL, Van Manen J, Vander Meer JJ. Demodulation of EMCs of pathological tremors. Development and testing of a demodulator for clinical use. Med Biol Eng Comp 1983; 21:17275. 14. Journ6e HL, Van Manen J. Improvement of the detectability of simulated pathological tremor EMGs by means of demodulation and spectral analysis. Med Biol Eng Comp 1983; 21:587-90. 15. Journ6e HL, Van Manen J.Automatic method to detect and determine the preferential frequency of EMC signals from patients with pathological tremors. Med Biol Eng Comp 1984; 22:32225. 16. Van Sittert NJ, Ribbens PH, Huisman B, et al. A nine-year followup study of renal effect in workers exposed to cadmium in a zinc ore refinery. Br J Ind Med 1993; 50:603-12. 17. Rosenman KD, Valciukas JA, Glickman, L, et al. Sensitive indicators of inorganic mercury toxicity. Arch Environ Health 1986; 41:20&15. 18. Singer R,Valciukas JA, Rosenman KD. Peripheral neurotoxicity in workers exposed to inorganic mercury compounds. Arch Environ Health 1987; 42:181-84. 19. Piikivi L, Ruokonen A. Renal function and long-term low mercury vapor exposure. Arch Environ Health 1989; 44:14649. 20. Ellingsen DG, BarregArd L, Carder PI, et al. Assessment of renal dysfunction in workers previously exposed to mercury vapour at a chloralkali plant. Br J Ind Med 1993; 50:881-87.

115