Neuromotor function in a cohort of Danish steel workers

NeuroToxicology 28 (2007) 336–344

Neuromotor function in a cohort of Danish steel workers Morten Blond *, Bo Netterstrom Hilleroed Hospital, Hilleroed, Denmark Received 20 October 2005; accepted 18 July 2006 Available online 2 August 2006

Abstract Objective: With a longitudinal design to evaluate possible neuromotor impairment in a cohort of steel workers exposed to metal dust. Material: Ninety-two employees from a steel works were examined in 1989 and 1995. Sixty were re-examined in 2003. A non-matched control group was examined in 1996 (n = 19) and in 2003 (n = 14). Median blood manganese in 1989, 1995 and 2003 was 149, 171 and 155 nmol/l. Median blood lead in 1989 and 2003 was 0.76 and 0.22 mmol/l. Median air concentration of manganese at the steel works was estimated to be 0.11 mg/m3 in 1970s and was 0.03 mg/m3 in 1990s. Median air concentration of lead was estimated to be 0.13 mg/m3 in 1970s and was 0.01 mg/m3 in 1990s. Method: The Catsys 2000TM system developed by Danish Product Development is computer-based device for measuring hand tremor, hand coordination and reaction time. Results: Over all there were no statistically significant differences in neuromotor function between the participating steel workers, nonparticipating steel workers and controls in 1995/1996. Only reaction time for the right hand was slower for the participating steel workers. Compared with the control group the steel workers showed a decline in the ability to perform fast precise hand pronation/supination and finger tapping from 1995 to 2005. Correlation analysis showed no associations between test results for fast hand coordination and blood manganese and lead. Only seniority was associated with deterioration of beat regulation of fast pronation/supination of the hands. Discussion: On a group basis the changes were subclinical, but they should none the less be taken seriously. Conclusion: Changes of neuromotor function measured as the ability to perform fast precise pronation/supination of the hands and fast precise finger tapping was shown in this cohort of steel workers. No causal relationships could be shown. # 2006 Elsevier Inc. All rights reserved. Keywords: Occupational exposure to manganese; Neuromotor; Steel worker; Catsys; Neurotoxicity

1. Introduction At our clinic in Denmark we have seen a number of patients with neurological complaints from a steel producing plant, where scrap metal was used as a raw material. They have suffered from neuromotor deficiencies, cognitive disturbances or both. Often exposure to manganese was considered the cause. Because of this, we have examined a cohort of long-term exposed steel workers from the plant with the objective of showing whether or not impairment of neuromotor function can be shown. Manganese (Mn) is a common metal and considered an essential micronutrient. Excessive occupational exposure can result in poisoning (manganism), first described in connection * Corresponding author. Tel.: +45 45422527; fax: +45 48294713. E-mail addresses: [email protected] (M. Blond), [email protected] (B. Netterstrom). 0161-813X/$ – see front matter # 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.neuro.2006.07.010

with mining (Couper, 1837). Subsequently, it has also been described in industry and agriculture (Ferraz et al., 1988). Traditionally, manganism is divided into three phases, based on Rodier’s description in 1955 of 150 cases of manganism. In the first phase, non-specific, subjective and neuropsychological symptoms are described, together with minor gait abnormalities. In the second phase, motor symptoms are the most prevalent, including impairment of speech and facial expressions, more severe gait impairment and adiadochokinesis. At the same time, the neuropsychological symptoms increase. The third, fully developed phase is characterized by rigidity of the muscles, pronounced gait impairment (pas du coq), tremor in the upper extremities, possible vegetative disturbances and deformities. Several authors have emphasized the similarity to Parkinson’s disease. During the last two decades, a number of epidemiological studies have been published, in which the authors have used different neuromotor tests to establish whether or not low to

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

moderate levels of manganese exposure lead to subclinical or minor neuromotor dysfunction. The results show a preponderance of positive results, but the findings are inconsistent. Most of the studies are cross-sectional in design. Only three studies are more or less of a longitudinal design (Crump and Rousseau, 1999; Lucchini et al., 1999; Roels et al., 1999). Some authors only account to a limited extent for negative test results and some perform a vast number of tests, of which only a few are positive. The control groups are seldom entirely comparable. Findings of dose–response relationships are inconsistent, both in and between studies. 2. Materials and methods 2.1. Materials In 1989 and 1995 two non-peer reviewed cross-sectional studies of 503 and 191 employees at the plant were published as internal reports (Hansen, 1989; Netterstrom and Raffn, 1996). It was suspected that the level of exposure to manganese in the electro-steel works, where scrap metal was melted in two large ovens, was detrimental to a persons’ health. Therefore the 1989 study consisted of 341 present and 37 former employees from the electro-steel works and as controls 125 employees from other parts of the plant. The 1995 study consisted of the employees from the electro-steel works most exposed to manganese, some employees on sick leave and 19 employees from other parts of the plant. In order to evaluate the long-term effect of working at a steel works, we designed the present study to include everyone who had participated in both of these two former studies, thereby selecting a group of employees with a certain minimum of seniority, with data from the previous studies and with supposed relatively high exposure. Ninety-two people had participated in both 1989 and 1995 and were included in the study in 2003. Sixty (65%) participated. Of these 25 (42%) held jobs, 17 were unemployed, 3 were on sick leave and 15 had retired. The plant closed down in the summer of 2002, but was subsequently partly reopened and started to

337

roll out sheet iron. A number of the 25 actively working participants were employed at the reopened plant at the time of examination in 2003, but unfortunately the number is unknown. In 2003 the melting ovens at the electro-steel works were not in use. As a consequence there was no longer exposure to vapours from melted steel and manganese was no longer handled in raw form. Exposure from dust and red-hot steel continued. All other participants had not been exposed to manganese for minimum half a year prior to examination. At the time of the study, nine participating steel workers had ongoing or previous workers’ compensation claims concerning manganism. Of the 32 nonparticipating steel workers, 4 were excluded for geographical reasons and 1 because of alcohol intoxication, while 27 did not with to participate in the study. At least nine non-participating steel workers had previously been examined for manganism in connection with workers’ compensation claims and a number of them stated this as the reason for not wanting further examination. Five employees have also worked at a nearby iron foundry for 1–11 years with a median of 3 years. These employments were included in the period of employment since manganese exposure is known to occur at iron foundries. The study design is shown in Fig. 1. A non-matched control group was used. Data regarding the control group were collected during an earlier study undertaken by the Clinic of Occupational Medicine at Hilleroed Hospital in 1996. The group consisted of 19 persons, who constituted the non-exposed control group in a study of printing workers exposed to toluene (Eller et al., 1999). They were employed as office workers, smiths and electricians. The control group was comparable to the participating group of steel workers regarding age (median 52 years, Mann–Whitney: p = 0.73), gender (14 males, Fischer’s exact test: p = 1.00) and smoking status 2003 (7 daily smokers, Fischer’s exact test: p = 0.77). Alcohol consumption in the control group was higher in 2003 (median 2.6 drinks/day, Mann–Whitney test: p = 0.04). Blood levels of manganese and lead for the control group were not measured, but assumed to be normal since there was no known occupational exposure for these metals.

Fig. 1. Study design.

338

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

Table 1 Background characteristics for the 92 manganese exposed steel workers Characteristic

Age at examination in 2003 Sex (number of males) Vocational training (number of unskilled) Current smoker 1995 (number of ‘‘Yes’’) Current smoker 2003 (number of ‘‘Yes’’) Alcohol 1995, drinks/dayc Alcohol 2003, drinks/dayc Blood manganese 1989 (nmol/l)d Blood manganese 1995 (nmol/l)d Blood manganese 2003 (nmol/l)d Blood lead 1989 (mmol/l)e Blood lead 2003 (mmol/l)e Years working in the most manganese exposed areas 1995 Total years employed at the plant 2003 a b c d e

Participants

Non-participants

P

n

Median

Range

n

Median

Range

60 60 53 56 60 50 60 60 52 60 49 60 58

56 57 22 34 25 0.8 1.7 149 171 155 0.76 0.23 8

35–68 – – – – 0.0–9.6 0.0–7.0 90–539 85–394 94–309 0.08–2.21 0.09–0.92 0–28

32 32 – 30 – 30 – 32 27 – 27 – 32

54 32 – 19 – 1.5 – 195 189 – 0.94 – 10

36–68 – – – – 0.0–7.0 – 114–466 136–270 – 0.24–2.22 – 0–22

7–44

–

–

–

60

24

0.73a 0.55b – 1.00b – 0.14a – 0.00a 0.01a – 0.15a – 0.69a –

Mann–Whitney test. Fishers exact test. One drink = 12 g alcohol. Reference values: median 157 nmol/l, 95% reference interval 100–271 nmol/l (Kristiansen et al., 1997). Reference values: median 0.24 mmol/l, 95% reference interval 0.11–0.53 mmol/l (Kristiansen et al., 1997).

Table 1 shows the background data of participating and nonparticipating steel workers from 1995. The table shows that participating and non-participating steel workers were comparable with regard to age, gender, smoking habits and exposure to manganese prior to 1995, based on the knowledge at that time concerning manganese exposure. Blood levels of manganese in the non-participating group were significantly higher in 1989 and 1995. A statistically non-significant trend in the nonparticipating group towards greater alcohol consumption and a higher blood level of lead was noted. 2.2. Methods The examination consisted of four elements: a questionnaire, blood sampling, Catsys 2000TM from Danish Product Development and Cognitive Function Scanner1. The Cognitive Function Scanner1 is a computer-based neuropsychological test battery, lasted 1 11/2 h and is the subject of a separate article. The Catsys 2000TM is a portable computer-based examination of tremor, diadochokinesis and simple reaction time. The examination lasted 20–30 min and was administered by one of the study physicians after the Cognitive Function Scanner1 test. The Catsys 2000TM standard test battery was used and the equipment was calibrated by the producer prior to testing. All tests were carried out separately for each hand. When measuring tremor the subject held a stylus horizontally 10 cm in front of his or her navel with a recording time of 8 s. The stylus is equipped with a biaxial accelerometer. The key output is the tremor intensity, defined as the root mean square of accelerations, recorded in the 0.9–15.0 Hz band and measured in m/s2. The rhythmic tests, alternating pronation and supination of the hand and finger tapping were obtained by slapping or tapping a drum following a constant metronome beat as

precisely as possible. Recording time was 20 s for slow 1.0 Hz tests and 10 s for fast 2.5 Hz tests. The key figures are precision or average offset, i.e. the average amount of time in seconds the hand or finger is from beating the drum on the metronome beats and the standard deviation of this precision, i.e. the beat regulation, also measured in seconds. The standard deviation is maybe the most important figure of the two, as it expresses the test subject’s ability to follow the rhythm, whether the drum is tapped a little to early or a little to late in relation to the metronome beat. During maximum frequency testing the subject followed a steadily increasing metronome frequency as well as possible. During the reaction time test 10 random combined visual and auditive signals were given during a 40 s test period. The test person held a handle and responded by pressing a switch with his or her thumb as quickly as possible. Average response time and standard deviation were recorded. The test system is described in detail elsewhere (Despre´s et al., 2000) and (www.catsys.dk). Statistical tests were carried out using SPSS 11.0 for Windows. Tests of normality were undertaken using Kolmogorov–Smirnov and Shapiro–Wilk tests. Because a large proportion of the results could not be assumed to be normally distributed, non-parametric tests were used. Two-sided p-values below 0.05 were considered to be statistically significant. The study was approved by the local scientific ethical committee and by the Danish Data Protection Agency and was carried out in accordance with the Helsinki 2 Declaration. 2.3. Exposure The plant produced recycled steel between 1942 and 2002. Over the years, extension and rebuilding has been ongoing. In 1982, the new electro-steel works was opened. In 1989, the steel production was approximately 600,000 tons and in 1995

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

approximately 750,000 tons. In 1995, 9000 tons of manganese was added to the steel in the form of ferromanganese and silicomanganese. The manganese arrived on ships from Norway and was handled both manually and mechanically. In the 1970s, measurements of total dust levels were made at the plant. The results varied from 0.7 to 62.6 mg/m3. Manganese typically constituted about 1–3% of the total dust concentration. Therefore the manganese level in the air in 1970s is assumed to have been between 0.01 and 1.9 mg/m3 with a median of 0.11 mg/m3. In the 1990s, the levels of manganese in the air were measured by stationary and personal methods and were between 0.01 and 0.84 mg/m3 in the electrosteel works with a median of 0.03 mg/m3. Two much higher values were measured, one stationary measurement on the top of a silo of 17.3 mg/m3 and another single stationary measurement between the furnaces of 4.1 mg/m3 at a time when production was halted due to repair and cleaning. In 2005, the Danish TLV for manganese in the air is 0.2 mg/m3. There are only few measurements of the levels of lead in the air in 1970s. Lead typically constituted about 0.5–4% of the total

339

dust concentration. Therefore the lead level in the air in 1970s is assumed to have been between 0.004 and 2.5 mg/m3 with a median of 0.13 mg/m3. In the 1990s, the levels of lead in the air were measured by stationary and personal methods and were between 0.001 and 1.72 mg/m3 in the electro-steel works with a median value of 0.01 mg/m3. The majority of measurements were below the former TLV of 0.1 mg/m3. In 2005, the Danish TLV for lead in the air is 0.05 mg/m3. Since 1990, employees at the steel works have worn ‘‘air-stream’’ helmets with P2 filters. Accordingly, a stationary measurement in 1997 showed 1.19 mg/ m3, while a personal measurement made inside a helmet showed 0.05 mg/m3 in the furnace area in the electro-works. However it has often been stated by former employees that the helmets were used with some carelessness and only in certain areas, so continued exposure must be suspected. Besides manganese and lead other metals have been measured a few times during the 1990s. Median nickel was 0.0018 mg/m3 (TLV 0.05 mg/m3), median zinc was 0.09 mg/m3 (TLV 4.0 mg/ m3), median cadmium was 0.00015 mg/m3 (TLV 0.01 mg/m3) and median iron was 0.28 mg/m3 (TLV 3.5 mg/m3).

Table 2 Catsys 2000TM variables (median values) 1995 Pa

Steel workers Participants (n = 56–60)

Non-participants (n = 29–32)

Controls Participants (n = 14)

Pb

Pc

Non-participants (n = 4–5)

Tremor intensity RH (m/s2) Tremor intensity LH (m/s2)

0.130 0.125

0.140 0.130

0.36 0.78

0.115 0.125

0.180 0.200

0.03 0.04

0.32 0.70

Slow Slow Slow Slow

0.061 0.071 0.045 0.044

0.057 0.058 0.042 0.042

0.98 0.48 0.28 0.68

0.074 0.070 0.040 0.040

0.104 0.111 0.052 0.038

0.64 0.58 0.06 0.61

0.67 0.38 0.13 0.83

0.026 0.012 0.021 0.023

0.022 0.011 0.018 0.021

0.20 0.48 0.35 0.52

0.026 0.019 0.022 0.022

0.031 0.060 0.083 0.048

0.33 0.10 0.06 0.13

0.53 0.99 0.96 0.79

0.032 0.027 0.038 0.036

0.026 0.026 0.039 0.035

0.51 0.33 0.89 0.89

0.037 0.030 0.039 0.037

0.057 0.055 0.039 0.036

0.46 0.25 0.85 0.93

0.77 0.77 0.92 0.64

0.015 0.018 0.015 0.017

0.015 0.014 0.016 0.017

0.96 0.43 0.80 0.77

0.024 0.025 0.017 0.017

0.016 0.025 0.021 0.018

0.85 0.64 0.03d 0.89

0.18 0.15 0.27 0.89

4.500 4.700 6.200 5.800

4.750 4.500 6.150 6.150

0.45 0.71 0.93 0.20

5.050 4.950 6.050 6.400

5.100 5.300 6.300 6.100

0.61 0.61 0.68 0.78

0.03 0.14 0.39 0.09

0.208 0.218 0.033 0.032

0.198 0.215 0.026 0.031

0.18 0.80 0.06 0.55

0.186 0.204 0.033 0.041

0.178 0.171 0.023 0.024

0.27 0.15 0.41 0.02

0.02d 0.19 0.80 0.35

Fast Fast Fast Fast

pron/sup pron/sup pron/sup pron/sup

Slow Slow Slow Slow Fast Fast Fast Fast

pron/sup pron/sup pron/sup pron/sup

finger finger finger finger

finger finger finger finger

Max. Max. Max. Max.

freq. freq. freq. freq.

Reaction Reaction Reaction Reaction

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

tapping tapping tapping tapping

tapping tapping tapping tapping

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

P/S RH (Hz) P/S LH (Hz) finger tapping RH (Hz) finger tapping LH (Hz)

time time time time

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

RH, right hand; LH, left hand; S.D., standard deviation; slow, 1 Hz; fast, 2.5 Hz; pron/sup, alternating pronation and supination of the hand; Max. freq., maximum frequency. a Participating steel workers 1995 vs. non participating steel workers 1995. Mann–Whitney test. b Participating controls 1995 vs. non-participating controls 1995. Mann–Whitney test. c Participating steel workers 1995 vs. participating controls 1995. Mann–Whitney test. d Indicates that the Kruskal–Wallis test performed on all four groups had an asymp. p-value < 0.05.

340

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

3. Results Table 2 shows the results of the Catsys tests from 1995. Only pair wise significant results marked with a d should be considered truly significant, since statistical testing of all four groups with the Kruskal–Wallis test was significant in these cases, thereby reducing the risk of error due to multiple comparisons of more than two groups. There was no indication of impaired neuromotor function among the non-participating steel workers compared with the participating steel workers. A tendency towards impaired neuromotor function was seen in the small non-participating part of the control group when compared to the participating part. Due to multiple comparisons only the result concerning S.D. of fast finger tapping with the right hand can be considered significant (column P Footnote b). The Catsys test results for the participating steel workers and the control group were comparable in 1995 (column P Footnote

c) with one exception. The reaction time for the right hand was statistically better in the control group. Table 3 shows the results of Catsys examination in this study. The columns P Footnote a–Footnote d display the statistical results of group comparisons. The cross-sectional analysis (column P Footnote a) showed significant differences between the participating steel workers and the control group in 2003. The steel workers displayed poorer regulation of slow and fast right hand pronation/supination and of fast left hand pronation/supination. As in 1995 the control group had a shorter right hand reaction time. Longitudinal analysis of the results for the steel workers from 1995 to 2003 (column P Footnote b) showed many statistically significant changes. Precision during right hand slow pronation/supination, precision and regulation during both maximum frequency pronation/supination and maximum frequency finger tapping with both hands and reaction time for the left hand all

Table 3 Catsys variables 1995 and 2003 (median values) for steel workers and controls Steel workers 1995 (n = 56–60) Tremor Tremor Tremor Tremor Tremor Tremor Slow Slow Slow Slow Fast Fast Fast Fast

pron/sup pron/sup pron/sup pron/sup

pron/sup pron/sup pron/sup pron/sup

Slow Slow Slow Slow Fast Fast Fast Fast

intensity RH (m/s2) intensity LH (m/s2) center frequency RH (Hz) center frequency LH (Hz) frequency S.D. RH (Hz) frequency S.D. LH (Hz)

finger finger finger finger

finger finger finger finger

Max. Max. Max. Max.

freq. freq. freq. freq.

Reaction Reaction Reaction Reaction

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

tapping tapping tapping tapping

tapping tapping tapping tapping

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

P/S RH (Hz) P/S LH (Hz) finger tapping RH (Hz) finger tapping LH (Hz)

time time time time

RH (s) LH (s) RH S.D. (s) LH S.D. (s)

Controls

Pa

Pb

Pc

Pd

2003 (n = 60)

1995 (n = 14)

2003 (n = 14)

0.130 0.125 – – – –

0.120 0.110 7.300 7.550 2.800 3.400

0.115 0.125 – – – –

0.100 0.120 7.200 7.450 2.900 3.300

0.11 0.90 0.82 0.78 0.53 0.56

0.20 0.16 – – – –

0.02 0.06 – – – –

0.26 0.43 – – – –

0.061 0.071 0.045 0.044

0.049 0.054 0.047 0.042

0.074 0.070 0.040 0.040

0.036 0.050 0.037 0.039

0.97 0.70 0.01 0.29

0.01 0.07 0.09 0.33

0.18 0.33 0.38 0.46

0.89 0.70 0.15 0.41

0.026 0.012 0.021 0.023

0.042 0.044 0.036 0.055

0.026 0.019 0.022 0.022

0.025 0.021 0.019 0.023

0.11 0.06 0.05 0.03

0.04 0.00 0.00 0.00

0.78 0.75 0.95 0.88

0.60 0.01 0.04 0.04

0.032 0.027 0.038 0.036

0.038 0.023 0.048 0.045

0.037 0.030 0.039 0.037

0.038 0.043 0.042 0.041

0.85 0.84 0.36 0.47

0.62 0.57 0.00 0.00

0.75 0.88 0.17 0.49

0.61 0.65 0.67 0.42

0.015 0.018 0.015 0.017

0.034 0.042 0.023 0.026

0.024 0.025 0.017 0.017

0.023 0.023 0.020 0.019

0.22 0.18 0.26 0.07

0.00 0.00 0.00 0.00

0.80 0.64 0.07 0.43

0.03 0.03 0.12 0.07

4.500 4.700 6.200 5.800

6.100 6.050 7.950 7.650

5.050 4.950 6.050 6.400

6.000 6.400 7.750 8.100

0.88 0.43 0.68 0.07

0.00 0.00 0.00 0.00

0.03 0.00 0.04 0.01

0.21 0.82 0.86 0.86

0.208 0.218 0.033 0.032

0.205 0.203 0.032 0.034

0.186 0.204 0.033 0.041

0.191 0.203 0.037 0.034

0.04 0.31 0.53 0.52

0.82 0.04 0.73 0.57

0.95 0.19 0.95 0.22

0.89 0.98 0.97 0.24

RH, right hand; LH, left hand; S.D., standard deviation; slow, 1 Hz; fast, 2.5 Hz; pron/sup, alternating pronation and supination of the hand; Max. freq., maximum frequency. a Steel workers 2003 vs. controls 2003. Mann–Whitney test. b Steel workers 1995 vs. steel workers 2003. Wilcoxon test. c Controls 1995 vs. controls 2003. Wilcoxon test. d Changes for steel workers from 1995 to 2003 vs. changes for controls from 1995 to 2003. Mann–Whitney test.

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

341

Table 4 Bivariate correlation analysis between the changes of fast rhythmic tests from 1995 to 2003 (N = 56–60) Measure of exposure

Spearman’s rho

Pron/sup RH

Pron/sup LH

Seniority

Pron/sup RH S.D.

Pron/sup LH S.D.

Correlation coefficient p (2 tailed)

0.13 0.31

Age

Correlation coefficient p (2 tailed)

Blood lead 2003

Correlation coefficient p (2 tailed)

Finger tapping RH

Finger tapping LH

Finger tapping RH S.D.

Finger tapping LH S.D.

0.06 0.67

0.28 0.04

0.54 0.00

0.04 0.78

0.06 0.63

0.02 0.91

0.20 0.15

0.08 0.57

0.10 0.46

0.27 0.05

0.21 0.12

0.14 0.28

0.06 0.65

0.19 0.15

0.18 0.20

0.04 0.74

0.01 0.93

0.23 0.09

0.22 0.10

0.10 0.47

0.16 0.23

0.01 0.93

0.08 0.55

Pron/sup, pronation/supination of the hand; RH, right hand; LH, left hand; S.D., standard deviation.

improved. Precision and regulation of fast pronation/supination with both hands and of fast finger tapping with both hands deteriorated. Regulation of slow finger tapping was also worse for both hands. Similar longitudinal analysis of the results for the control group (column P Footnote c) also showed improvement of all four maximum frequency tests. Tremor intensity of the right hand increased. Comparisons of the changes from 1995 to 2003 revealed a pattern (column P Footnote d). The changes for the steel workers during fast pronation/supination and fast finger tapping were significantly different from that of the controls in five tests, showed trends in two tests and was non-significant in the last of eight test parameters. The changes express deterioration in ability. The changes for all other tests showed no statistical differences between the two groups. In order to examine the association between exposures and neuromotor function, bivariate correlation analysis was performed. No significant associations were shown with fast rhythmic Catsys test variables as dependent and blood lead and manganese as independent variables. A trend was found between blood manganese in 1995 and the S.D. of fast right hand pronation/supination (Spearman’s rho 0.26 with p = 0.06). The statistical test results were more consistent for an association between seniority and the S.D. of fast left hand pronation/supination (Spearman’s rho 0.42 with p = 0.001). Multivariate linear regression analysis incorporating age did not alter this association. However the adjusted R squared was 0.14 indicating, as would be expected, that the association only partially explains the results of the Catsys test. Table 4 shows the statistically significant and nearly significant results of bivariate correlation analysis between measures of exposure and the changes in Catsys test results from 1995 to 2003. The results regarding blood manganese in 1995 and 2003, blood lead in 1989 and 1995 and gammaglutamyl transferase are not shown. An association between the S.D.-changes of both fast pronation/supination tests and seniority was seen. Age was associated with S.D.-change of right hand pronation/supination. There was a trend towards association with blood lead. The statistically significant correlation coefficients all represent a decline in ability to perform the rhythmic tests. Multivariate linear regression analysis incorporating age shows the same association with seniority, while age is not significantly associated.

4. Discussion and conclusions In newer epidemiological studies of manganese exposure some authors have found hand tremor increased (Bast-Pettersen et al., 2004; Hochberg et al., 1996; Mergler et al., 1994; Roels et al., 1987, 1992) and some have found changes in tremor characteristics in the form of a higher centre frequency and reduced dispersion without increased tremor intensity, changes that resemble a pathological tremor (Beuter et al., 1999; Lucchini et al., 1999). The present study adds itself to the list of negative studies in this respect. The ability to perform alternating movements with the fingers has been examined and found compromised in some studies (Chia et al., 1993; Iregren, 1990; Lucchini et al., 1995; Myers et al., 2003b; Sjogren et al., 1996), but not in all (BastPettersen et al., 2004; Gibbs et al., 1999; Hua and Huang, 1991; Lucchini et al., 1999; Mergler et al., 1994). Dose–response association between this ability and measures of manganese exposure have been found (Lucchini et al., 1999; Myers et al., 2003b), but most often not (Chia et al., 1993; Gibbs et al., 1999; Iregren, 1990; Mergler et al., 1999; Myers et al., 2003a; Sjogren et al., 1996). Despite low exposure levels at the steel works an effect on finger tapping was shown in this study. The ability to perform alternating movements with the hands or the arms has also been examined and found compromised in some studies (Bast-Pettersen et al., 2004; Lucchini et al., 1999; Mergler et al., 1999), but not in all (Beuter et al., 1994, 1999; Sjogren et al., 1996). There is conflicting information from one study (Wennberg et al., 1991, 1992), however at least a trend towards compromised ability was shown. Dose–response association between the ability to perform alternating hand or arm movements and measures of manganese exposure has been found in one study (Myers et al., 2003b), but most often not (Bast-Pettersen et al., 2004; Myers et al., 2003a; Sjogren et al., 1996). This study finds an effect, but as most other studies no dose–response relationship was shown. Besides tests of finger tapping and alternating pronation and supination of the hands many studies have incorporated dexterity tests, in which an individuals ability to perform fast, coordinated movements with his or her hands and fingers is measured. These tests have shown both reduced ability (Chia et al., 1993; Hochberg et al., 1996; Mergler et al., 1994, 1999; Myers et al., 2003b; Roels et al., 1987, 1992; Sjogren et al.,

342

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

1996) and not reduced ability (Bast-Pettersen et al., 2004; Gibbs et al., 1999; Hua and Huang, 1991; Lucchini et al., 1997; Wennberg et al., 1991). Dose–response association has been found by some authors (Beuter et al., 1999; Lucchini et al., 1997; Roels et al., 1987, 1992), but not by others (Chia et al., 1993; Myers et al., 2003b; Sjogren et al., 1996; Wennberg et al., 1991). Simple reaction time has been found to be prolonged or with a greater dispersion by some authors (Roels et al., 1987, 1992; Siegl and Bergert, 1982). In the present study 1 of 4 test results showed a difference between the two groups, but this difference is due to the control group having a fast reaction time for the right hand. No deterioration in reaction time was seen for the steel workers, no dose–response effect was found and the results are consistent with no effect. The results are interpreted as indicating a decline of ability to perform rapid and accurate movements with the hands and fingers. This interpretation is in accordance with one of the common – but inconsistent – findings reported in literature: impaired function or association with manganese exposure when testing with Purdue pegboard test, Santa Ana pegboard test, pursuit aiming test, eurythmokinesimetri or orthokinesimetri (Beuter et al., 1999; Chia et al., 1993; Lucchini et al., 1997; Mergler et al., 1999; Myers et al., 2003b; Roels et al., 1987, 1999; Sjogren et al., 1996). The three longitudinal studies of manganese exposure have not shown clear patterns of impairment of the ability to perform fast and precise movements with the hands and fingers. Crump reported in 1999 the results of 11 years of monitoring at a manganese oxide and salt producing plant. A subgroup consisted of the workers Roels described in 1987. Eye-hand coordination, measured with an orthokinesimeter, changed towards both better and worse results during the first years of monitoring, where after the results stabilized. All in all there was no evidence that the neurological changes observed by Roels progressed towards clinically detectable signs. In 1999 Lucchini published a casecontrol study of 61 cases, of which 30 had been examined with the finger tapping test of the Swedish Performance Evaluations System in 1991 and 1997. No statistical differences between the results of the two measurements were found. However no analysis of the 31 dropouts was reported. An improved working environment was stated as a possible explanation. In 1999 Roels published a follow-up study of 92 individuals examined in 1987. Thirty-four present and 24 former employees were re-examined. Among the 34 employees improvement of eye-hand coordination associated to a decline in exposure to manganese was reported. Similar results were obtained among the former employees. As expected, the blood levels of lead in 1989 were relatively high, metal works being one of the most exposed industries in Denmark (Rasmussen and Gilkou, 1991). In a cohort study from the greater Copenhagen area, a decrease in the median blood level of lead of 45% from 0.62 to 0.34 mmol/l was found in males during the 11-year period from 1976 to 1987 (Christensen and Holst, 1988). In a newer study, the mean concentration of lead in a Danish reference population was 0.17 mmol/l (Nielsen et al., 1998). In the present study, a

decrease of 70% was seen during a 14-year period. This reduction can be explained by the widespread use of unleaded gasoline during recent years, improved working conditions at the plant and by the fact that quite a few had left the plant. The median level of 0.76 mmol/l in 1989 was lower than the new Danish limit of 0.96 mmol/l for continued surveillance, and the highest levels were lower than the limit of 2.9 mmol/l which, according to the Danish Working Environment Authority in 2001 requires a medical examination. Some studies investigating the effect of lead on manual dexterity have shown significant associations (Hanninen et al., 1998; Jeyaratnam et al., 1986; Mantere et al., 1982; Ryan et al., 1987; Schwartz et al., 1993; Schwatrz et al., 2005). A recent summary of two meta-analyses also showed impairment of psychomotor function (Seeber et al., 2002). However both the manganese and the blood lead levels in the present study were low in comparison with previously reported values. When doing many statistical tests one must be aware of the risk of type 1 error. In principle 1 out of every 20 performed statistical tests would be significant at an alfa level of 0.05 merely due to chance. If a full Bonferroni-adjustment was made to an overall alfa of 0.05 hardly any significant results would have been found. Such a correction would however have been too conservative since the test results are correlated. Even after a Bonferroni-adjustment corrected for the mean correlation coefficients found during analysis hardly any significant results would have been found. However the goal of such an adjustment is to keep the type 1 error to a minimum, i.e. the finding of 5% or more significant statistical test results by chance—but not of finding a pattern of several significant results. Occurrence of significant results by chance would be expected to be randomly distributed, which is clearly not the case when comparing the changes in Catsys variables between the steel workers and the controls. In this study the authors argue that the finding of increased deterioration of the test results during fast rhythmic tests represents such a pattern. Examiner bias can result in an unevenly distributed pattern of test results. The pronounced findings of increased maximum frequencies for both groups are most likely the result of such a bias. By comparing changes over time the influence of examiner bias is reduced. In accordance with this there are no differences between the changes in maximum frequencies over time for the two groups. On the other hand the pattern of changes in performance during fast rhythmic testing is not eliminated by the analysis and therefore most likely not the result of examiner bias. The higher level of manganese in the blood among the nonparticipating steel workers may be attributable to an increased probability of workers with high blood levels having been examined in connection with a worker’s compensation claim and therefore not wanting to participate in the study. Such a possible selection bias would however not weaken positive findings in the study and cannot be conceived as the explanation of a pattern of impaired ability to perform fast rhythmic movements. The non-participating controls performed somewhat worse in 1996 than the participating controls. Had these five

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

individuals participated there is a possibility that comparison of the Catsys test results in 2003 would have resulted in fewer significant results being observed and this can be considered a possible weakness of the study. Age and lifestyle factors such as alcohol and tobacco use are potential confounders. There were no differences between the groups regarding age and tobacco use. The control group reported a higher alcohol consumption, which does not weaken positive findings. Less marked changes would be expected to occur in the control group, since the time span between the examinations was shorter, i.e. 7 years instead of 8 years. Again such a difference cannot explain the finding of a pattern. The results of our correlation analysis show that the study’s measures of exposure including blood levels of manganese and lead seemingly are inadequate. It can therefore by hypothesised that our findings are the result of combined exposures or other exposures. Over 20 years of occupational exposure to the combination of lead and iron has been shown to be associated with an increased odds ratio of having Parkinson’s disease in one study (Gorell et al., 1999). Cadmium has been found associated with peripheral neuropathy (Hart et al., 1989; Viaene et al., 1999). Also physical workload including hand-arm vibrations could be a potential causative exposure. Manual dexterity has been shown to be reduced in patients with handarm vibration syndrome (Toibana et al., 2002). Other exposures would also suggest an explanation for the apparent paradox that manual dexterity declines at the same time that environmental exposures decrease. Studies stating reference values for the Catsys test battery have been published. One set of published reference values for tremor intensity, precision, rhythmic regulation and reaction time was published after testing 150 people (Despre´s et al., 2000). These values are comparable to the values of the participating steel workers in 2003. A second set of values for precision, rhythmic regulation and reaction time is published after testing 76 women with a median age of 37 years (Gyntelberg et al., 1990). There is a tendency towards the steel workers’ values for fast pronation/supination of the hand being worse in comparison with these younger women while a similar pattern is less obvious for finger tapping. All median values are situated well within the reported 95% percentiles. The Catsys 2000TM manual states a set of reference values. Compared to these the same tendency towards compromised fast pronation/supination of the hand is seen among the steel workers. On a group basis the steel workers’ impaired ability to perform rapid and accurate movements with the hands and fingers must be seen as subclinical and not as an expression of clinically apparent disease. However since even a smaller shift of group performance can substantially increase the number of persons falling outside reference intervals, even small decreases in group performance should be taken seriously. In conclusion the results of this study are consistent with sub clinical changes of the neuromotor abilities of the examined long term employed steel workers, making fast and accurate movements with the hands and fingers more difficult. Overall no dose–response effect was shown for the used measures of exposure for which reason it is not possible to attribute these observations to any particular exposure including exposure to

343

metals. Therefore new longitudinal studies preferably with greater contrasts in exposure are required. Acknowledgments The authors would like to thank the participants in the study as well as the staff at the Clinic of Occupational Medicine at Hilleroed Hospital for their efforts. A special thank you goes to Doctor Keld Johnsen, who participated in the planning and implementation of the study. Financial support was received from the research funds administered by the Research Council for Health Services in Frederiksborg County. Statistical support was given by the Department of Biostatistics, University of Copenhagen. References Bast-Pettersen R, Ellingsen DG, Hetland SM, Thomassen Y. Neuropsychological function in manganese alloy plant workers. Int Arch Occup Environ Health 2004;77(4):277–87. Beuter A, Mergler D, de Geoffroy A, Carriere L, Belanger S, Varghese L, et al. Diadochokinesimetry: a study of patients with Parkinson’s disease and manganese exposed workers. Neurotoxicology 1994;15(3):655–64. Beuter A, Edwards R, deGeoffroy A, Mergler D, Hundnell K. Quantification of neuromotor function for detection of the effects of manganese. Neurotoxicology 1999;20(2/3):355–66. Chia SE, Foo SC, Gan SL, Jeyaratnam J, Tian CS. Neurobehavioral functions among workers exposed to manganese ore. Scand J Work Environ Health 1993;19(4):264–70. Christensen JM, Holst E. Evaluation of blood lead levels in Danes for the period 1976–1987. Anal Bioanal Chem 1988;332(6):710–3. Couper J. On the effects of black oxide of manganese when inhaled into the lungs. Br Ann Med Pharm 1837;1:41–2. Crump KS, Rousseau P. Results from eleven years of neurological health surveillance at a manganese oxide and salt producing plant. Neurotoxicology 1999;20(2/3):273–86. Despre´s C, Lamoureux D, Beuter A. Standardization of a neuromotor test battery: the CATSYS system. Neurotoxicology 2000;21(5):725–35. Eller N, Netterstrom B, Laursen P. Risk of chronic effects on the central nervous system at low toluene exposure. Occup Med (Lond) 1999;49(6):389–95. Ferraz HB, Bertolucci PH, Pereira JS, Lima JG, Andrade LA. Chronic exposure to the fungicide maneb may produce symptoms and signs of CNS manganese intoxication. Neurology 1988;38(4):550–3. Gibbs JP, Crump KS, Houck DP, Warren PA, Mosley WS. Focused medical surveillance: a search for subclinical movement disorders in a cohort of U.S. workers exposed to low levels of manganese dust. Neurotoxicology 1999;20 (2/3):299–313. Gorell JM, Johnson CC, Rybicki BA, Peterson EL, Kortsha GX, Brown GG, et al. Occupational exposure to manganese, copper, lead, iron, mercury and zinc and the risk of Parkinson’s disease. Neurotoxicology 1999;20(2/ 3):239–47. Gyntelberg F, Flarup M, Mikkelsen S, Palm T, Ryom C, Suadicani P. Computerized coordination ability testing. Acta Neurol Scand 1990;82(1):39–42. Hanninen H, Aitio A, Kovala T, Luukkonen R, Matikainen E, Mannelin T, et al. Occupational exposure to lead and neuropsychological dysfunction. Occup Environ Med 1998;55(3):202–9. Occupational Exposure and Health Effects in the Danish Steel Industry. Department of Occupational Medicine, Rigshospitalet, State University Hospital, Copenhagen, Denmark and The National Institute of Occupational Health; 1989. Hart RP, Rose CS, Hamer RM. Neuropsychological effects of occupational exposure to cadmium. J Clin Exp Neuropsychol 1989;11(6):933–43. Hochberg F, Miller G, Valenzuela R, McNelis S, Crump KS, Covington T, et al. Late motor deficits of Chilean manganese miners: a blinded control study. Neurology 1996;47(3):788–95.

344

M. Blond, B. Netterstrom / NeuroToxicology 28 (2007) 336–344

Hua MS, Huang CC. Chronic occupational exposure to manganese and neurobehavioral function. J Clin Exp Neuropsychol 1991;13(4):495–507. Iregren A. Psychological test performance in foundry workers exposed to low levels of manganese. Neurotoxicol Teratol 1990;12(6):673–5. Jeyaratnam J, Boey KW, Ong CN, Chia CB, Phoon WO. Neuropsychological studies on lead workers in Singapore. Br J Ind Med 1986;43(9):626–9. Kristiansen J, Christensen JM, Iversen BS, Sabbioni E. Toxic trace element reference levels in blood and urine: influence of gender and lifestyle factors. Sci Total Environ 1997;204(2):147–60. Lucchini R, Selis L, Folli D, Apostoli P, Mutti A, Vanoni O, et al. Neurobehavioral effects of manganese in workers from a ferroalloy plant after temporary cessation of exposure. Scand J Work Environ Health 1995;21(2):143–9. Lucchini R, Bergamaschi E, Smargiassi A, Festa D, Apostoli P. Motor function, olfactory threshold, and haematological indices in manganese-exposed ferroalloy workers. Environ Res 1997;73(1/2):175–80 [Erratum in: Environ Res 1997 Nov;75(2):187]. Lucchini R, Apostoli P, Perrone C, Placidi D, Albini E, Migliorati P, et al. Longterm exposure to ‘‘low levels’’ of manganese oxides and neurofunctional changes in ferroalloy workers. Neurotoxicology 1999;20(2/3):287–97. Mantere P, Hanninen H, Hernberg S. Subclinical neurotoxic lead effects: twoyear follow-up studies with psychological test methods. Neurobehav Toxicol Teratol 1982;4(6):725–7. Mergler D, Huel G, Bowler R, Iregren A, Belanger S, Baldwin M, et al. Nervous system dysfunction among workers with long-term exposure to manganese. Environ Res 1994;64(2):151–80. Mergler D, Baldwin M, Belanger S, Larribe F, Beuter A, Bowler R, et al. Manganese neurotoxicity, a continuum of dysfunction: results from a community based study. Neurotoxicology 1999;20(2/3):327–42. Myers JE, teWaterNaude J, Fourie M, Zogoe HB, Naik I, Theodorou P, et al. Nervous system effects of occupational manganese exposure on South African manganese mineworkers. Neurotoxicology 2003a;24(4/5):649–56. Myers JE, Thompson ML, Ramushu S, Young T, Jeebhay MF, London L, et al. The nervous system effects of occupational exposure on workers in a South African manganese smelter. Neurotoxicology 2003b;24(6):885–94. Netterstrom B, Raffn E. A cross-sectional study of manganese exposed workers at The Danish Steelworks (authors translation). Report no. 6. Denmark: Department of Occupational Health, Hillerod Hospital, 1996. Nielsen JB, Grandjean P, Jorgensen PJ. Blood lead concentration in the Danish population after introduction of lead-free gasoline. Ugeskr Laeger 1998;160 (33):4768–71. Rasmussen K, Gilkou T. Health surveillance of workers exposed to lead in the County of Arhus. Ugeskr Laeger 1991;153(14):975–8.

Rodier J. Manganese poisoning in Moroccan miners. Br J Ind Med 1955;12 (1):21–35. Roels H, Lauwerys R, Buchet JP, Genet P, Sarhan MJ, Hanotiau I, et al. Epidemiological survey among workers exposed to manganese: effects on lung, central nervous system, and some biological indices. Am J Ind Med 1987;11(3):307–27. Roels HA, Ghyselen P, Buchet JP, Ceulemans E, Lauwerys RR. Assessment of the permissible exposure level to manganese in workers exposed to manganese dioxide dust. Br J Ind Med 1992;49(1):25–34. Roels HA, Ortega Eslava MI, Ceulemans E, Robert A, Lison D. Prospective study on the reversibility of neurobehavioral effects in workers exposed to manganese dioxide. Neurotoxicology 1999;20(2/3):255– 271. Ryan CM, Morrow L, Parkinson D, Bromet E. Low level lead exposure and neuropsychological functioning in blue collar males. Int J Neurosci 1987;36(1/2):29–39. Schwartz BS, Bolla KI, Stewart W, Ford DP, Agnew J, Frumkin H. Decrements in neurobehavioral performance associated with mixed exposure to organic and inorganic lead. Am J Epidemiol 1993;137(9):1006–21. Schwatrz BS, Lee BK, Bandeen-Roche K, Stewart W, Bolla K, Links J, et al. Occupational lead exposure and longitudinal decline in neurobehavioral test scores. Epidemiology 2005;16(1):106–13. Seeber A, Meyer-Baron M, Schaper M. A summary of two meta-analyses on neurobehavioural effects due to occupational lead exposure. Arch Toxicol 2002;76(3):137–45. Siegl P, Bergert KD. A method of early diagnostic monitoring in manganese exposure. Z Gesamte Hyg 1982;28(8):524–6. Sjogren B, Iregren A, Frech W, Hagman M, Johansson L, Tesarz M, et al. Effects on the nervous system among welders exposed to aluminium and manganese. Occup Environ Med 1996;53(1):32–40. Toibana N, Ishikawa N, Sakakibara H. Measurement of manipulative dexterity in patients with hand-arm vibration syndrome. Int Arch Occup Environ Health 2002;75(1/2):106–10. Viaene MK, Roels HA, Leenders J, De Groof M, Swerts LJ, Lison D, et al. Cadmium: a possible etiological factor in peripheral polyneuropathy. Neurotoxicology 1999;20(1):7–16. Wennberg A, Iregren A, Struwe G, Cizinsky G, Hagman M, Johansson L. Manganese exposure in steel smelters a health hazard to the nervous system. Scand J Work Environ Health 1991;17(4):255–62. Wennberg A, Hagman M, Johansson L. Preclinical neurophysiological signs of parkinsonism in occupational manganese exposure. Neurotoxicology 1992; 13(1):271–4.