Escolar Documentos
Profissional Documentos
Cultura Documentos
www.occenvmed.com
104 Viaene, Pauwels, Veulemans, et al
microsomal epoxide hydrolase (mEH) en- family income (people who were economically
zyme.9 In humans, mEH shows polymorphic less important to their family—for example,
genotypes at amino acid positions 113 and 139 being married to a working spouse and having
that substantially alter mEH activity between no children—were preferentially dismissed).
the diVerent isoenzymes.10 Studies of the Three years after the event in 1989, we
enzyme activity of the mEH phenotype showed retraced the 185 workers enrolled in our previ-
a twofold to 63-fold variation that may partly ous studies.16 17 Figure 1 shows how the popu-
be due to methodological diVerences.11–14 lation of the present retrospective study was
Competitive inhibition of mEH has been recruited and found to be eligible on the basis
reported to cause an excess of central nervous of initial defined exclusion criteria (craniocer-
system toxicity15 suggesting that mEH with dif- ebral trauma with 10 minutes or more uncon-
ferent activities may have important neurotoxi- sciousness, alcoholism (five or more alcoholic
cological consequences associated with expo- drinks/day),18 epilepsy, major depression, or
sure to styrene. other major neurological disease). In total, 117
The present investigation is a retrospective subjects of the original workforce participated
cohort study in workers exposed to styrene in the present study (participation rate 63%)
from a polyester boat building plant who consisting of 90 formerly exposed workers and
constituted one historical cohort that experi- 27 workers currently exposed to low concen-
enced the same exposure situation until the trations of styrene. A group of 111 control male
usual industrial activity stopped in 1989. Most subjects (electricians and assembly workers)
of the workforce was dismissed. They formed from another plant were invited to cooperate in
the basis for the formerly exposed group of the the study. Age, sex, and schooling level were
present study. A limited number of workers matched as closely as possible across groups.
continued to work in repair and maintenance Twenty seven refused to participate and 20
of styrene polyester boats but their exposure to were excluded according to initial defined
styrene dropped by a factor of about four at the exclusion criteria (the same criteria as those
time of the present study. Almost all of them applied to the exposed subjects as well as the
constituted the currently exposed group. absence of a history of occupational exposure
The present epidemiological study was to solvents). Eventually, 64 control workers
carried out in the winter of 1992–3. Its first aim (58%) were enrolled in the study. The
was to compare on a group basis possible neu- neurobehavioural testing of these groups oc-
rotoxic symptoms and neurobehavioural ef- curred from November 1992 to March 1993.
fects in formerly and currently exposed work- Unexpectedly, in February 1994 the plant
ers with a group of control workers who had with the workers still employed at low exposure
never been occupationally exposed to organic planned to stop activities. A strike period of 1
solvents or other neurotoxic chemicals. The year followed and eventually the plant shut
second aim was to assess in the combined down in February 1995. These events caused a
exposed group (currently and formerly ex- delay in blood sampling for characterisation of
posed workers) possible associations between the mEH phenotype. Blood was taken from
neurotoxic eVects and historical industrial September 1994 to June 1995—that is, about
hygiene data on exposure to airborn sty- 6–15 months after the end of exposure in cur-
rene,16 17 such as concentration, duration of rently exposed workers and 5–6 years after the
exposure, or cumulative exposure. The third formerly exposed workers left the plant. In
aim of the study was to investigate whether total, 56 workers (48%) from the combined
there may exist a link between mEH phenotype group of currently and formerly exposed work-
activity and neurobehavioural performance in ers were willing to provide a blood sample.
workers exposed to styrene. The study also had
the potential to investigate whether neurotoxic CHARACTERISATION OF EXPOSURE TO STYRENE
eVects linked to styrene were persistent 3 years Despite a well designed ventilation system,
after the end of or a substantial decrease in 4%–9% of the personal air samples exceeded
moderate to high occupational exposure to sty- the previous American Conference of Govern-
rene. ment Industrial Hygienists (ACGIH) thresh-
old limit value time weighted average (TLV-
Subjects and methods TWA)19 of 213 mg/m3 and 4% of the peak
STUDY POPULATION exposures at the most exposed job sites
The production of polyester boats started early exceeded the previous TLV-STEL (short time
1982. From 1984 to 1987, we carried out sev- exposure limit)19 of 426 mg/m3 as shown in the
eral cross sectional industrial hygiene studies in earlier industrial hygiene studies from 1984 to
the styrene polyester boat building plant with a 1987.16 17 Detailed individual records were kept
workforce that at that time consisted of 185 for every worker as to the diVerent tasks during
exposed male employees.16 17 In 1989, the usual his daily working hours (job characteristics,
production activities stopped for economic duration, and location of every individual task)
reasons and only repair and maintenance allowing calculation of the exact total number
activities were continued with a reduced work- of hours every individual worker was eVectively
force (n=31) that could continue to work at the exposed to styrene from 1982 to 1993 for cur-
plant. Due to the changed industrial activities, rently and from 1982 to 1989 for formerly
the exposure to styrene dropped considerably exposed workers. Also, instant measurements
to an average of 41 mg/m3. The other workers with colorimetric Dräger tubes were carried
(n=154) were made redundant on the basis of out daily by the industrial hygienist at locations
age (workers more than 50 years old retired) or where exposure to styrene was suspected to be
www.occenvmed.com
Styrene neurotoxicity 105
www.occenvmed.com
106 Viaene, Pauwels, Veulemans, et al
was blinded to the intensity and duration of exposed workers one way analysis of variance
exposure of the workers. (ANOVA) and Duncan’s multiple range test
were used for continuous outcome variables
ANALYSIS OF mEH PHENOTYPE ACTIVITY (number of complaints and NES test results).
BuVy coat was collected from two blood General ANOVA models were built allowing
samples of 8 ml, each withdrawn by venepunc- for age, alcohol consumption, years of welding,
ture with Vacutainer CPTTM cell preparation and the number of years at school as
tubes (Becton-Dickinson USA) containing covariables for the NES results and addition-
sodium citrate as anticoagulant. The mononu- ally smoking and personality score for the
clear cells were separated and rinsed twice with questionnaire results. General ANOVA results
phosphate buVered saline and then S-TKM are presented with å2 values (proportion of
buVer (0.25 M sucrose, 80 mM Tris, 25 mM variance in the dependent variable explained
KCl, 5 mM MgCl2, 0.1 mM EDTA, pH 7.35) by the diVerence across the groups). Multiple
was added to obtain a 2 ml suspension which linear regression or multiple logistic regression
was sonicated on ice followed by microcen- analyses (dichotomous variables) were applied
trifugation at 9000g at 4°C. Aliquots of the S9 on the combined exposed group (currently and
fraction were taken for protein measurement formerly exposed workers) to study dose-eVect
by the BIORAD method,23 and typical protein or response relations. The regression models
concentrations were between 1000 and 2000 were run with the forced entry method by
µg/ml buVer solution. The S9 fraction solutions introducing one of the exposure indices at the
were stored at −80°C for further use. time (duration of exposure, mean concentra-
A novel mEH assay has been developed in tion of exposure, interaction term between
which a labelled styrene epoxide was used as both, or cumulative exposure) together with
substrate. Briefly 5 µl of a 10-4 M styrene glycol the dependent variables and covariables. For
solution in isotonic saline was added to 240 µl the SNAC results and the NSC-60 category
of the S9 fraction as internal standard. The scores, age, alcohol consumption, the number
reaction was initiated by addition of 5 µl of years at school, years of welding, smoking,
[2H8]styrene-7,8-oxide (10-4 M in acetone) and and personality score were the covariables. In
incubation for 30 minutes in an oil bath at multiple regression models of the NES results,
37°C. The incubation mixture was extracted age, alcohol consumption, the number of years
with 300 µl ethyl acetate; 100 µl of the ethyl at school, and years of welding were included
acetate layer was dried under an N2 steam. The as independent covariables. Belonging to the
mixture was derivatised for 30 minutes at current exposure group or not (1 or 0) was also
110°C with BSTFA/acetonitril (1/10), and included as a dummy variable in the regression
analysed by GC-MS. The derivatised styrene analyses next to the exposure indices and the
glycol samples were injected by means of a covariables to test separately the eVect of being
solid injector (kept at 253°C) into an HP 5890 currently exposed. For the eVect of mEH phe-
series-II gas chromatograph coupled with an notype activity on the outcome variables, alco-
HP 5970 quadrupole mass spectrometer. For hol consumption and age were included as
chromatographic separation of the compounds covariables. The level of significance was set at
of interest, a DB-5-ms fused silica capillary a p value <0.05.
column (30 m×0.25 mm, 0.12 µm phase thick-
ness, Alltech Associates, USA) was used with a Results
source pressure of 0.006 Pa, the interface was CHARACTERISTICS OF THE STUDY GROUPS
kept at 270ºC and the oven programmed at The participating groups did not diVer greatly
7ºC/minute from 100°C–140°C and at 30°C/ for the covariables of interest (table 1). Despite
minute from 140°C–300°C. The mass spec- eVorts to match the groups according to
trometer operated in the electron impact mode schooling level, the currently and formerly
with an ionisation energy of 70 eV. By referring exposed workers had fewer years of schooling
to a calibration curve, the monitoring of and thus also a lower schooling level than the
ion-185 (trimethylsilated deuterated styrene controls. It is interesting to note that in the
glycol) relative to ion-174 (undeuterated ana- three groups the scores for motivation of
logue, internal standard) allowed an accurate participation in the study were equally high (>
measurement of the mEH mediated formation 98%) and that their personality scores were
of styrene glycol in the incubated samples. The similar. The currently exposed workers who
mEH phenotype activity was assessed as the laminated polyester from 1982 to 1992, on
mean of triplicate incubations. average were exposed for 4700 hours, whereas
the formerly exposed workers laminated only
STATISTICAL ANALYSIS from 1982 to 1989 (about 1100 hours less).
The statistical analysis was performed with the The overall mean concentration of exposure to
statistical package for social sciences (SPSS) styrene during their lamination tasks from
for Windows, version 7.0. When needed log 1982 to 1989 amounted to 148 and 157 mg/m3
normal transformation was applied (number of for currently and formerly exposed workers
SNAC complaints, symbol-digit substitution, respectively, the combined exposed workers on
simple reaction time). Fisher’s exact test (dou- average had 155 mg/m3 of exposure. Thus,
ble sided) and odds ratios (ORs) were calcu- before the 1989 event the exposure to styrene
lated for dichotomous outcome variables—for did not diVer significantly between currently
example, persistent SNAC complaints and and formerly exposed workers. This was also
NSC-60 categories. For comparisons between shown by the time weighted exposure concen-
control workers, and currently and formerly tration.
www.occenvmed.com
Styrene neurotoxicity 107
Table 1 Characteristics (mean (SD, range)) of the exposed and control groups
Age (y) 37.1 (7.9, 27–51) 38.1 (10.3, 26–64) 40.3 (11.3, 25–63)
Alcohol consumption (n drinks/week) 8.6 (8.6, 0–30) 8.7 (8.3, 0–35) 9.7 (10.2, 0–35)
Sleeping (n hours) 7.1 (0.8, 5–8) 7.1 (0.9, 5–9) 7.5 (0.8, 5–9)
Smoking (n cigarettes/day) 5.0 (7.5, 0–20) 6.9 (9.2, 0–30) 3.9 (7.2, 0–25)
Motivation (%) 98.3 (5.4, 75–100) 98.1 (6.8, 50–100) 99.3 (3.6, 80–100)
Personality score (NSC-60)* 1.50 (0.54, 1.00–3.00) 1.50 (0.44, 1.00–2.57) 1.53 (0.48,
1.00–3.29)
Years of schooling 11 (2, 7–14) 11 (2, 7–21) 13 (3, 7–24)
Schooling level:
Primary 22% 13% 5%
First grade 63% 67% 61%
Second grade or above 15% 20% 34%
Duration of exposure (hours of lamination)† 4700 (1870, 530–7505) 3610 (2000, 165–8098) —
Overall concentration of exposure to styrene
while laminating (mg/m3)‡ 148 (14.1, 124–173) 157 (19.0, 99–189) —
Time weighted exposure (mg/m3)§ 68 (31.1, 6–129) 70 (43.9, 1–174) —
*Neurotoxicity symptom checklist-6021: score between 1 (never complains) and 4 (always complains).
†Referring to period 1982–92 for the currently exposed group and to period 1982–89 for formerly exposed group.
‡Mean calculated from the average exposure concentration of each person at lamination tasks for the period 1982–9 (for details, see
section subjects and methods).
§Mean calculated from the individual cumulative exposure concentrations of styrene divided by the total numbers of working hours
in the plant. Period 1982–9 for both exposed groups.
The available characteristics for the workers exposure to styrene condition after 1989
exposed to styrene who did not participate had (mainly due to a new type of boat introduced at
a similar age distribution (mean (SD, range) the beginning of 1992), some complaints were
39.0 (10.6, 26–65)), but their educational level still overreported—such as fatigue in the
was lower than in participating currently and evening, common colds, and sleeping diYcul-
formerly exposed workers (primary 34%, first ties.
grade 58%, second grade or higher 6%). The By comparison with control workers, of
mean duration of exposure (3230 h) and over- which 29 workers (45%) ever experienced
all mean exposure concentration (146 mg/m3) some complaint during work, significantly
were not diVerent from those of participating more formerly exposed workers (79%) experi-
exposed workers. enced some kind of complaint during lamina-
tion tasks versus only 52% in currently exposed
QUESTIONNAIRE RESULTS workers (table 3). Although both currently and
The results for each single complaint question formerly exposed workers had worked under
of the SNAC questionnaire are summarised in the same exposure conditions until 1989, there
table 2. During the moderate to high exposure was a significantly higher proportion of cur-
to styrene, the combined exposed group had a rently exposed workers (48%) who never expe-
significantly higher frequency of complaints rienced complaints during lamination than
such as irritability, fatigue in the evening, head- formerly exposed workers (21%) (Fisher’s
ache, common colds, intolerance for alcohol, exact test, p<0.01; OR 3.5; 95% confidence
diminished endurance, memory, and smell interval (95% CI) 1.4–8.6). The number of
than did the controls. Despite not having been complaints per person during exposure was
exposed for 3 years at the time of the SNAC twice as high in currently exposed and four
questionnaire, 10 of the formerly exposed sub- times as high in formerly exposed workers
jects still complained of a persistently dimin- compared with controls (one way ANOVA,
ished sense of smell (p<0.01). In the currently p<0.01). The excess of complaints in the
exposed workers, who experienced lowered exposed groups remained unchanged after
Table 2 Structured neurological anamnesis of complaints (SNAC): results (n (%)) on single complaint questions
www.occenvmed.com
108 Viaene, Pauwels, Veulemans, et al
Table 3 Structured neurological anamnesis of complaints (SNAC): sum of all complaint questions
**p<0.01.
†At the time of anamnesis (1992–3) after shut down of original production activities in 1989: the currently exposed group had 3
years of low exposure to styrene, and the formerly exposed group had ended exposure 3 years previously. Until 1989, both groups
had the same moderate to high exposure to styrene.
‡Unadjusted results=raw data; adjusted results are the raw data adjusted for age, schooling, alcohol use, years of welding, smoking,
and personality score.
§p<0.05 v controls; ¶p<0.05 v the two other groups, Duncan’s multiple range test.
allowing for covariables in the model—such as and digit span forwards (p<0.01). Hand-eye
age, schooling, alcohol consumption, years of coordination was significantly (p<0.01) worse
welding, smoking, and personality score (AN- only in formerly exposed workers compared
COVA: exposure group, å2=0.18, p<0.01; none with control workers. After allowing for cov-
of the covariables reached significance). For ariables (age, alcohol consumption, number of
the period after the 1989 event, low exposure years at school, and years of welding) the group
for currently exposed workers and 3 years after eVect for symbol-digit substitution, digit span
the end of exposure for formerly exposed forwards, and hand-eye coordination remained
workers, the proportion of formerly exposed unchanged (ANCOVA: å2=0.05, p=0.01;
workers still reporting complaints markedly å2=0.09, p<0.01; å2=0.05, p<0.01, respec-
decreased from 79% to 31%. The mean tively). Age was a significant covariable for the
number of complaints per person in formerly three neurobehavioural tests (symbol-digit
exposed workers drastically decreased as well substitution, å2=0.32; digit span forwards,
and did not diVer from that in controls, å2=0.06; hand-eye coordination, å2=0.06; all
whereas currently exposed workers still re- p<0.01) and, also, the years of schooling
ported more complaints than the two other explained a significant part of the symbol-digit
groups (one way ANOVA, p=0.07). Currently substitution variance (å2=0.05, p<0.01). Sim-
exposed workers still showed an excess number ple reaction time, colour-word vigilance, asso-
of complaints after allowing for the covariables ciated learning, and associated recall did not
in the model (ANCOVA: exposure group, diVer significantly between the groups.
å2=0.03, p=0.07; none of the covariables
reached significance). ACTIVITY OF THE mEH PHENOTYPE
For the NSC-60 questionnaire, significantly The assay for the activity of mEH phenotype
more formerly exposed workers exceeded the had a coeYcient of variation of 13% for three
cut oV scores of several categories: chest repeated measurements in the same person.
(formerly exposed 36 (40%) v control workers Enzyme assay blanks with incubation of heat
14 (22%), p=0.02); equilibrium (formerly denatured S9-fraction showed a non-
exposed 28 (31%) v control workers 10 (16%), enzymatic conversion of 3%–4% of the sub-
p=0.04); and somatic complaints (formerly strate into [2H8]styrene glycol. The mEH phe-
exposed 66 (73%) v control workers 34 (53%), notype activity was normally distributed in the
p<0.01). In currently exposed workers only the studied population and ranged from 0.54 to
equilibrium category score reached borderline 2.68 pmol/mg protein/min (mean (SD) 1.57
significance (currently exposed nine (33%) v (0.49)) which represented a fivefold variation
control workers 10 (16%), p=0.09). The in mEH phenotype activity between people.
proportions of workers that exceeded the cut
oV scores for the chest and somatic complaints DOSE-EFFECT OR DOSE-RESPONSE RELATIONS
categories were similar in currently and for- In multiple linear regression, the number of
merly exposed workers, nine (33%) v 36 (40%) complaints per person in currently and formerly
and 18 (67%) v 66 (73%) respectively. exposed workers during the period of high
exposure (thus before the 1989 event) was
NES TEST RESULTS significantly related to the individual overall
In general, currently and formerly exposed mean concentration of exposure to styrene
workers performed worse than control workers (regression coeYcient â=0.25, p=0.03), but not
(table 4). One way ANOVA and Duncan’s to duration of exposure (â=0.09, p=0.34) or
multiple range test showed that compared with cumulative exposure (â=0.08, p=0.44). The
control workers currently and formerly ex- interaction term concentration of
posed workers had significantly worse test exposure×duration of exposure did not reach
results for symbol-digit substitution (p=0.01) significance (â=0.13, p=0.19). None of the cov-
www.occenvmed.com
Styrene neurotoxicity 109
p Value
>0.10
>0.10
0.02
0.06
−0.16 0.09
0.03
−0.17 0.06
personality score, smoking, alcohol use) was sig-
Interaction
nificantly associated with the number of com-
0.18
plaints (all p>0.2). Neither exposure indices,
0.05
0.17
0.21
0.02
â
nor any of the covariables explained the persist-
††Multiple linear regression analysis. Regression coeYcients (â) and p values for the predictor variables, duration of exposure, cumulative exposure, and the interaction term duration×concentration of exposure are presented.
ence of complaints at the time of the interview in
p Value
>0.10
>0.10
>0.10
0.05
−0.17 0.08
0.07
−0.18 0.05
formerly exposed workers (multiple logistic
Cumulative
0.05
0.14
0.18
0.02
ariables in the model, the dummy variable
Dose-eVect relation††
<0.01
>0.10
>0.10
>0.10
showed a significant association with the
−0.16 0.08
0.03
−0.18 0.05
Duration of
0.08
0.18
0.21
0.04
plaints than being formerly exposed (â=−0.15,
â
−3.5
−1.3
0.1
(27.9, 20 to 180)
(74, 449 to 742)
6.9 (2.1, 2 to 9)
Controls (n=64)
558
−3.5
−1.3
(25.3, 17 to 150)
558
†p<0.05 v controls, ‡ p<0.05 v other two groups, ANOVA one way Duncan’s multiple range test.
(0.9,† −2.26 to 1.07)
results**
adjusted
p value
>0.10
<0.01
>0.10
>0.01
>0.10
0.01
www.occenvmed.com
110 Viaene, Pauwels, Veulemans, et al
10
0.1
5
0.0
0
–0.1
–5
–0.2 –10
0 1500 3000 4500 6000 7500 9000 0.5 1.0 1.5 2.0 2.5 3.0
Duration of exposure Microsomal epoxide hydrolase activity
(total number of hours laminating) (pmol/mg protein/min)
Figure 2 Linear regression model in the exposed study Figure 3 Linear regression model in the mEH phenotype
population (+, currently exposed and ï, formerly exposed) study group (n=56) between the mEH phenotype activity
(n=117) between the duration of exposure and symbol-digit and symbol-digit substitution test results after adjustment
substitution test results after adjustment for age, educational for age, alcohol use, educational level, and duration of
level, alcohol consumption, and years of welding (residual exposure (residual log (symbol-digit substitution)).
log (symbol-digit substitution)).
ibility of neurobehavioural eVects.2 Although
with controls. The estimated value for the diYcult to carry out, the ideal study design in
upper limit of normal for symbol-digit substi- this context is a longitudinal follow up after the
tution time is 3.3 s/digit (mean of the controls end of exposure in a relatively large cohort of
+2SDs), which is similar to normal data workers. Economic and trade union linked
reported by Dutch investigators (3.2 s/digit).24 instabilities of the late 1980s forced us to
This suggests that under the given exposure change our investigation into a design aiming at
conditions, the evolution from a normal mean collecting retrospectively neurological infor-
(2.5 s/digit) to reach pathological values of the mation in a cohort of workers exposed to
symbol-digit substitution time will be reached styrene in which 83% lost their job while the
on average after 15 160 hours of lamination. rest continued to work in the same plant at job
This is equivalent to a full time exposure of 8.6 sites gradually decreasing in exposure to
years. A similar calculation based on the styrene. In those field conditions, the maxi-
relation between symbol-digit substitution mum that could be done to answer the crucial
time and the interaction term, duration of question of whether neurotoxic complaints
exposure×concentration of exposure, showed were persistent in this workforce was to organ-
that a pathological value for symbol-digit sub- ise a structured neurological anamnesis of
stitution time would be reached at an interac- complaints in such a way, that reliable
tion product of 2 419 220 mg/m3×hour or responses could be obtained from the partici-
1375 mg/m3×year (taking 220 working days a pants 3 years after the 1989 event in the plant.
year and 8 hours work a day). This means 6.5 We are confident that a potential bias did not
years of full time lamination tasks at the previ- operate in the recruitment and selection proce-
ous styrene TLV-TWA value of 213 mg/m3 or dure of the subjects in the present study. Sixty
9.2 full time working years at a mean exposure three per cent of the original cohort of workers
to styrene concentration of 155 mg/m3. Al- exposed to styrene participated in the study.
though the safety measures and the ventilation For non-participant workers, all basic group
system in the plant were very well designed and and exposure characteristics were available and
resulted in low to moderate concentrations of were similar to the study group, except for a
exposure to styrene, it should be pointed out lower educational level. This could have
that up to 4% of the air measurements were lowered the robustness of the study, as it has
above the previous ACGIH TLV-STEL and up been suggested that low intellectual capacity
to 9% of the air measurements were above the may predispose to the development of neuro-
previous TLV-TWA depending on location and toxic eVects of organic solvents.8 It can be
characteristics of the lamination task.16 17 argued that the division of the original
In multiple regression analysis, no relation workforce into currently and formerly exposed
was found between the number of complaints workers could have introduced a bias due to
per person and the mEH phenotype activity the procedure by which workers were made
(â=0.17, p=0.25), nor with the persistence of redundant. However, no medical or productiv-
complaints (p=0.59). There was a significant ity criteria were used in this procedure. The
relation between the mEH phenotype activity criterion of age (>50 years) is not likely to have
and symbol-digit substitution test results introduced a bias as there was no relation
(â=0.34, p=0.01, fig 3) or colour-word vigi- between age and duration of exposure in this
lance (â=0.28, p=0.05), and a borderline study population. Older workers were exposed
significant relation with digit span forwards to lower air concentrations of styrene and they
(â=−0.24, p=0.08). The test results worsened had spent less hours on lamination tasks. Simi-
with increasing activity of the mEH phenotype. larly, years of welding, years or level of
education, and alcohol use were not associated
Discussion with concentration or duration of exposure.
As far as we know, only one short term follow In line with complaints described in other
up study about occupational exposure to studies,1 2 the present study also showed a
styrene has dealt with the question of revers- significantly higher prevalence of complaints
www.occenvmed.com
Styrene neurotoxicity 111
(irritability, fatigue, headache, alcohol intoler- and chronic eVects, due to the integrated dose.
ance, concentration diYculties, imbalance, and This has also been suggested by others.25
diminished sense of smell) when workers were The neurobehavioural test findings are in
exposed to styrene compared with control line with most other styrene studies reporting
workers. The ambient air concentration of eVects on attention demanding tasks, visuomo-
styrene during lamination tasks was the most tor performance, learning, and memory.3 4 26 In
important factor explaining the number of the present study, hand-eye coordination (vis-
SNAC complaints at work. Interestingly, 3 uomotor accuracy), digit span forwards (short
years after the end of exposure (mean TWA 70 term memory, attention), and symbol-digit
mg/m3 before the 1989 event), formerly ex- substitution (visuomotor performance) test
posed workers still reported an excess of atypi- results were significantly diVerent between for-
cal SNAC and NSC-60 (chest, equilibrium, merly or currently exposed workers and
somatic) complaints. These included epigastric controls. Unlike many who consider reaction
discomfort, headache, diYculties in swallow- time tests to be the most sensitive tests, the lack
ing, heart palpitations, dyspnoea, nausea, tinni- of eVect on simple reaction time in the workers
tus, diminished sense of smell, imbalance, chest currently exposed to very low concentratios of
oppression, feelings of drunkenness, and in- styrene is consistent with findings of Edling et
creased sensitivity to cold.21 In long term but al2 and supports the view that slowing of reac-
less exposed workers (mean TWA 36 mg/m3), tion time reflects acute narcotic eVects of
Edling et al2 reported that most complaints were higher concentrations of exposure.27 28
reversible after the end of exposure to styrene. Symbol-digit substitution time showed sig-
In the present study, none of the predictors such nificant relations with the duration of exposure
as concentration of exposure, duration of expo- and the interaction term, duration of
sure, activity of the mEH phenotype, or any exposure×exposure concentration, whereas
other covariable could explain the persistence digit span forwards showed a trend to worsen
of complaints at the time of the interview. as exposure time increased. Although no group
Unlike those of formerly exposed workers, the diVerences could be found, reaction times of
NSC-60 and SNAC results of currently ex- colour-word vigilance (perceptual speed) and
posed workers did not equally reach signifi- associated learning tests worsened significantly
cance, most probably because of a smaller with increasing duration of exposure in the
group size and because of a significantly greater exposed group. None of these variables was
proportion of currently exposed workers that related to the overall mean concentration of
never reported complaints although their expo- exposure. The results suggested that less than
sures did not diVer from those formerly 10 years of exposure to a mean atmospheric
exposed. This confirms the view that long term styrene concentration of 155 mg/m3 (STEL
exposed workers may experience less subjective measurements) or 70 mg/m3 (TWA measure-
complaints than workers who left exposure ear- ments) and the concomitant occurrence of up
lier. Alternatively, the excess of complaints in to 9% and up to 4% of these measurements
formerly exposed workers might be seen as being above the previous TLV-TWA and TLV-
reflecting a revenge attitude of the dismissed STEL values, resulted in the development of
workers. Although this could explain part of the dose-dependent eVects and the persistence of
variance, it is likely to be less plausible as the some complaints and neurobehavioural
number of complaints at the time the workers changes. This would indicate that sporadic
were exposed was well associated with the con- peak exposures may play a part in the develop-
centration of exposure. It should also be ment of chronic or persistent neurobehavioural
pointed out that at the time of the study all the eVects. This is in line with publications on
formerly exposed workers had been hired by organic encephalopathies due to exposure to
other companies and were thus no longer solvents,29 and strengthens the recent lowering
unemployed. Also, formerly exposed workers of the TLV-TWA and TLV-STEL concentra-
had the same results on the hold questions of tions by ACGIH from 213 and 416 mg/m3 to
the NSC-60 (personality category) as currently 85 and 170 mg/m3, respectively.30
exposed and control workers. This is a strong This study also showed a fivefold variation of
indication that formerly exposed workers had the activity of the mEH phenotype between
no tendency to tick high scores on the people, a variation in line with some of the
questionnaire items. An interviewer bias for the activity ranges previously reported, ninefold,11
SNAC results seems unlikely, not only because fivefold,12 and 1.6-fold.13 There was no associ-
of the blinding of the interviewer, but also ation between activity of the mEH phenotype
because a comparison between the results of the and age, alcohol use, overall mean concentra-
SNAC questionnaire and the self administered tion of exposure, or duration of exposure (all
NSC-60 questionnaire proved the consistency p>0.10). Activity of the mEH phenotype did
of the results.20 Despite lower TWA exposure at not seem to influence either the number of
the time of the study (40 mg/m3) a slightly complaints during exposure or the persistence
higher total number of complaints per person of the complaints. By contrast, visuomotor
remained in currently exposed workers (dimin- performance (symbol-digit substitution) and
ished sense of smell, increased fatigue in the perceptual speed (colour-word vigilance) were
evening, imbalance, and sleeping diYculties at worse in fast styrene metabolisers, but only if
night) but a dose-eVect relation could not be the duration of exposure was included in the
found. This excess of complaints in currently multiple regression analysis. In other words,
exposed workers could very well be a mixture of although exposure to styrene is still the primary
eVects, due to the current day to day exposure, factor determining the development of neu-
www.occenvmed.com
112 Viaene, Pauwels, Veulemans, et al
robehavioural eVects, the occurrence of these 6 Cherry N, Gautrin D. Neurotoxic eVects of styrene: further
evidence. Br J Ind Med 1990;47:29–37.
eVects may be modulated by the mEH pheno- 7 Dick RB. Neurobehavioral assessment of occupationally rel-
type. A similar pattern was found for the role of evant solvents and chemicals in humans. In: Chang LW,
Dyer RS, eds. Handbook of neurotoxicology. New York: Mar-
glutathione S-transferase M1 null genotype in cel Dekker, 1995:217–323.
the development of organic psychosyndrome, 8 Bleecker ML, Bolla KI, Agnew J, et al. Dose-related
in which the eVect became only significant in subclinical neurobehavioral eVects of chronic exposure to
low levels of organic solvents. Am J Ind Med 1991;19:715–
highly exposed workers.31 The slowing of 28.
visuomotor speed (increase of symbol-digit 9 International Agency for Research on Cancer. Some
industrial chemicals. IARC monographs on the evaluation of
substitution time) with increasing activity of carcinogenic risks to humans. Vol 60. Geneva: World Health
the mEH phenotype (fig 3) was surprising and Organisation, 1994.
10 Hassett C, Aicher L, Sidhu JS, et al. Human microsomal
might have diVerent explanations. Firstly, it epoxide hydrolase: genetic polymorphism and functional
could be that styrene-7,8-oxide is not the most expression in vitro of amino acid variants. Hum Mol Genet
1994:3:421–8.
toxic metabolite of styrene in humans, but 11 Glatt HR, Halfer-Wirkus H, Herborn J, et al. Interindividual
another toxic metabolite further down the variations in epoxide detoxifying enzymes. In: Müller JH,
Weber W, eds. Familial cancer. Basel: Karger, 1985:242–7.
metabolic chain could be. Mutti et al32 33 12 Nakajiama T, Elovaara E, Gonzalez FP, et al. Characteriza-
suggested that the ultimate metabolite, phe- tion of human cytochrome P450 isozymes responsible for
nylglyoxylic acid, could be the active neurotoxi- styrene metabolism. In: Abstract book of the International
Symposium on Health Hazards of Butadiene and Styrene.
cant. A second explanation might be that Helsinki: Finnish Institute of Occupational Health, 1993:
glutathione S-transferase activity in humans is 23.
13 Kroetz DL, Kerr BM, McFarland LV, et al. Measurement of
more important for styrene detoxification in in vivo microsomal epoxide hydrolase activity in white sub-
non-hepatic tissues than is commonly thought. jects. Clin Pharmacol Ther 1993;53:306–15.
14 Omiecinski CJ, Aicher L, Holubkov R, et al. Human periph-
On the other hand, the lack of glutathione eral lymphocytes as indicators of microsomal epoxide
S-transferase M1 could possibly lead to an hydrolase activity in liver and lung. Pharmacogenetics 1993;
3:150–8.
aspecific induction of other metabolic path- 15 Beiswanger CM, Mandella RD, Greassle TR, et al.
ways, for instance activity of the mEH Synergistic neurotoxic eVects of styrene oxide and
phenotype,34 which may indirectly contribute acrylamide: glutathione-independent necrosis of cerebellar
granule cells. Toxicol Appl Pharmacol 1993;118:233–44.
to a positive relation between activity of the 16 Veulemans H, Daems J, Masschelein R. (Polyester ship
mEH phenotype and neurotoxic eVects. building. Air monitoring and biomonitoring.) Cahiers
Arbeidsgeneeskunde 1986;22:75–83. (In Dutch.)
17 Daems J, Masschelein R, Veulemans H. (Health care in a
Conclusions fibreglass boat building factory.) Cahiers Arbeidsgeneeskunde
1986;23:289–295. (In Dutch.)
Some subjective symptoms or complaints and 18 British Special Committee of the Royal College of Psychia-
neurobehavioural eVects in workers exposed to trists. Alcohol, our favorite drug. London: Tavistock Publica-
tions, 1986.
styrene were most likely persistent, even at 19 American Conference of Governmental Industrial Hygien-
TWA exposure concentrations which were ists. Threshold limit values for chemical substances, physical
agents and biological exposure indices. Cincinnati: ACGIH,
thought to be safe in the 1980s. The concentra- Technical AVairs OYce, 1992.
tion of styrene in the air during lamination 20 Viaene MK. (Acute and chronic eVects of styrene exposure
and their persistence. Results of a cross-sectional study of
tasks seemed to be the most important factor in workers working in polyester ship-building industry) [PhD
the development of complaints, although most thesis]. Leuven, Belgium: Department of Occupational
Health, Katholieke Universiteit Leuven, 1998. (In Dutch.)
of these complaints disappeared after exposure 21 Hooisma J, Emmen HH. (Defining the normal data of the
ended. Also, the duration of exposure at neurotoxicity symptom checklist-60 (NSC-60)). Amsterdam:
lamination tasks seemed to be the best predic- Publicatie Stichting Arbouw, 1992. (In Dutch.)
22 Letz R. Use of computerized test batteries for quantifying
tor of toxic eVects on visuomotor performance neurobehavioral outcomes. Environ Health Perspect 1991;
and perceptual speed, which could still be 90:195–8.
23 Bradford MM. A rapid and sensitive method for quantifica-
detected in workers 3 years after the exposure tion of µg quantities of protein utilizing the principle of
to styrene had ended. The detoxification protein-dye binding. Anal Biochem 1976;72:248–54.
24 van der Laan G, van Dun RECS, Roos Y, et al. Organic psy-
metabolism of styrene-7,8-oxide may also chosyndrome caused by solvents? A protocol for diagnosis.
influence the development of these dysfunc- Amsterdam: Nederlands Centrum voor Beroepsziekten.
Coronel Laboratorium, Universiteit Amsterdam, 1995. (In
tions, although no clear conclusion about the Dutch.)
underlying mechanism can be drawn from the 25 Gregersen P. Neurotoxic eVects of organic solvents in
exposed workers: two controlled follow up studies after 5.5
present study. The results suggested that less and 10.6 years. Am J Ind Med 1988;14:681–701.
than 10 years of exposure to a mean concentra- 26 Anger WK. Worksite behavioral research: results, sensitive
methods, test batteries and transition from laboratory data
tion of atmospheric styrene of 155 mg/m3 may to human health. Neurotoxicology 1990;11:629–720.
result in persistent neurotoxic eVects. 27 Schoenhuber R, Gentilini M. Influence of occupational sty-
rene exposure on memory and attention. Neurotoxicol Tera-
tol 1989;11:585–6.
We are particularly indebted to the participants in the study, the 28 Mackay CJ, Kelman GR. Choice reaction time in workers
plant management and the occupational hygienist for their col- exposed to styrene vapour. Hum Toxicol 1986;5:85–9.
laboration and technical assistance. We gratefully thank Etienne 29 Houck P, Nebel D, Milham jr S. Organic solvent
Philips for his help in supervising tests and data entry. This encephalopathy: an old hazard revisited. Am J Ind Med
project was supported by the research programme of the Fund 1992;22:109–15.
for Scientific Research (Flanders) NFWO, Project No 30 American Conference of Governmental Industrial Hygien-
3.0175.96. ists. Threshold limit values for chemical substances, physical
agents and biological exposure indices. Cincinnati: ACGIH,
1 Flodin U, Ekberg K, Andersson L. Neuropsychiatric eVects Technical AVairs OYce, 1997.
of low exposure to styrene. Br J Ind Med 1989:46:805–8. 31 Söderkvist P, Ahmadi A, Åkerbäck A, et al. Glutathione
2 Edling C, Anundi H, Johanson G, et al. Increase in S-transferase M1 null genotype as a risk modifier for
neuropsychiatric symptoms after occupational exposure to solvent-induced chronic toxic encephalopathy. Scand J
low levels of styrene. Br J Ind Med 1993;50:843–50. Work Environ Health 1996;22:360–3.
3 Letz R, Mahoney FC, Hershman DL, et al. Neurobehavioral 32 Mutti A, Falzoi M, Romanelli A, et al. Regional alterations
eVects of acute styrene exposure in fibreglass boat builders. of brain catecholamines by styrene exposure in rabbits.
Neurotoxicol Teratol 1990;12:665–8. Arch Toxicol 1984;55:173–7.
4 Mutti A, Mazzucchi A, Rustichelli P, et al. Exposure-eVect 33 Mutti A, Falzoi M, Romanelli A, et al. Brain dopamine as a
and exposure-response relationships between occupational target for solvent toxicity: eVects of some monocyclic
exposure to styrene and neuropsychological functions. Am aromatic hydrocarbons. Toxicology 1988;49:77–82.
J Ind Med 1984;5:275–86. 34 Petruzzelli S, Camus AM, Carrozzi L, et al. Long-lasting
5 Yuasa J, Kishi R, Eguchi T, et al. Study of urinary mandelic eVects of tobacco smoking on pulmonary drug metaboliz-
acid concentration and peripheral nerve conduction among ing enzymes: a case-control study on lung cancer patients.
styrene workers. Am J Ind Med 1996;30:41–7. Cancer Res 1988;48:4695–700.
www.occenvmed.com