Abstract

Metalworking fluids (MWF) may contain a variety of additives, including medium-chain chlorinated paraffin (MCCP). There is concern that exposure to MCCP may be associated with increased risks for kidney toxicity. MCCP has been undergoing an evaluation as part of a European regulatory risk assessment and consequently it is important that an estimate is made of the likely occupational exposure. This paper reviews the available published data on dermal exposure to MWF and derives typical and reasonable worst-case (RWC) estimates of MCCP exposure. Estimates were highest for oil-based fluids (520 and 35 000 mg typical daily exposure and RWC exposure, respectively) and lowest for water-based fluids (30 and 520 mg for typical and RWC, respectively). Comparison with published inhalation exposure data suggests dermal and inhalation exposure may be of comparable magnitude.

INTRODUCTION

Metal working fluids (MWF) are complex mixtures of chemicals that are used in a range of engineering and metal processing applications. There are many MWF formulation types, from straight oils such as mineral oils to water-based fluids, which include soluble oils and synthetic fluids. These substances can pose a risk to health by inhalation of MWF aerosol with consequent airway toxicity (for example, Eisen et al., 2001; Robertson et al., 2007) and from dermal contact with the MWF producing local effects on the skin (Sprince et al., 1996). Exposure to straight oil MWF has also been linked to increased risks of bladder cancer (Friesen et al., 2009). In addition, exposure by either inhalation or dermal contact may give rise to systemic uptake of some of the chemicals contained in the mix.

MWF also contain a variety of additives and contaminants, including substances designed to prevent bacterial growth and/or improve odor, plus contaminants such as metal particles (also known as fines) and biological agents that may build up in the fluid as it is used.

Medium-chain chlorinated paraffin (MCCP) is used in MWF as an extreme pressure additive. MCCP has been considered as part of a European-wide risk assessment process under the European Union (EU) Existing Substances Regulations and so it is important to have information on exposure, particularly in relation to systemic exposure. The main health concern for this substance, based on animal data, is kidney toxicity (SCHER, 2008).

MCCP generally have a chlorine content of between 40 and 60% by weight and are manufactured by chlorination of liquid paraffin or paraffin that has a chain length of C14–17. The molecular weight for the MCCP products used in MWF is generally between 460 and 490.

Not all MWF contains MCCP. However, water-based MWF formulations may typically contain 5–10% MCCP, with chlorine content of 45–56%. Some formulations can contain up to 20% MCCP but this is usually only possible with lower viscosity MWFs. These products are diluted with water substantially reducing the in-use concentration of MCCP. The recommended dilution is ∼5% aqueous emulsion, i.e. one part MWF to 20 parts water. The estimated in-use concentration of MCCP in water-based fluid should therefore on average be ∼0.4% (data provided by EuroChlor).

Oil-based or neat fluids are not diluted by the end user and the chlorinated paraffin content range in these products is very wide, with some MWF products containing no MCCP and others as high as 100% MCCP, i.e. neat MCCP that is used for heavy duty applications such as broaching. However, for the main uses of oil-based fluids containing MCCP the content is 5–10%. The usage of this type of MWF differs between European countries. In mainland Europe, and particularly for Germany, chlorine-free chemistry is used for all low to medium duty machining applications. However, some heavy-duty applications use products with higher levels of MCCP, typically 50–70%. In the UK, the concentration of MCCP in oil-based MWF is generally lower than this range.

The main aim of the present work was to review the available literature on dermal exposure to MWF taking into account technical information on the use of MCCP in MWF, which was mainly obtained from the trade association representing companies manufacturing MCCP (EuroChlor), and derive estimates of dermal exposure to MCCP. In addition, we compare the dermal exposure estimates with published information on inhalation exposure to evaluate the relative importance of both routes.

METHODS

This paper presents dermal MWF exposure data obtained from three published sources, Roff et al. (2004); van Wendel de Joode et al. (2005); and Semple et al. (2007), which are the key publications with quantitative information about dermal MWF exposure. The data from these studies are reviewed in the following sections.

For the MCCP exposure estimation, information on the range and average MCCP content in MWF (as a percentage by mass) was used along with the dermal exposure data to estimate the typical (i.e. the median of the data) and a reasonable worst-case (RWC; i.e. 90th percentile) dermal exposure to MCCP in MWF. These results are expressed as milligram MCCP dermal exposure per day. Because the estimates of dermal exposure to MCCP in MWF were obtained from surrogate data we also made allowance for associated uncertainties.

Dermal exposure data were generally only available for hand exposure. However, we reviewed the available dermal MWF exposure data for non-hand exposure to estimate the likely exposure on other parts of the body.

No exposure estimates were made for inhalation exposure, although clearly this route may contribute to total systemic exposure. The results from the dermal exposure estimates are discussed in relation to published data for MWF inhalation exposure.

RESULTS

Study 1 (Semple et al.)—summary of findings

The authors of this study collected data on MWF exposure from a small sample of engineering companies in Scotland in 2004. This was an intervention study to assess the effectiveness of different safety training approaches. Six engineering firms were recruited via links with the trade association, a search of the Internet, and from professional contacts of the researchers. The firms were not intended to be representative of the entire engineering sector but covered a wide range of product and service delivery, including reconditioning large items of military equipment through the manufacturer of electric motors. Five of the six sites employed >100 workers and one site employed <50 people, although often the numbers involved in the machine tool departments were much lower and the number of workers directly exposed to MWF was generally between 10 and 40.

Gloves were not worn routinely by most machine operators due to the perceived risk of entanglement in moving machinery. This presented difficulty as the authors had intended to measure dermal exposure using interception-sampling techniques with cotton sampling gloves. Instead, a modified version of a wrist sampler was used (Vermeulen et al., 2000). Initial use of this patch method resulted in very low or zero dermal exposure levels being recorded. However, observation of workers showed that there was visible dermal exposure to the hands, but almost no collection of fluid on the wrist patch. Data from the patches were not presented in this paper as they were considered to underestimate exposure.

Wipe sampling was then selected as a more appropriate methodology to measure dermal exposure of the hands. Trials for two sampling materials were conducted to determine the number of wipes and wipe timings required to achieve suitable recovery efficiency, as assessed by wiping a known mass of MWF applied to the skin of a volunteer. The first material tested was a dry wipe of 10 × 10 cm comprising 60% cotton and 40% polyester cloth. The second wipe material was a similar sized moist hand wipe. Recovery was tested by applying a known volume of water-mix MWF directly on to the surface of the wipe material or on to the surface of a hand. After a specified period, the hand was wiped once on the palm, once on the back of the hand, and once on the length of each finger to collect any MWF residue.

All samples were analyzed by Inductively Coupled Plasma/Atomic Emission Spectrometry using boron as a marker of MWF contamination and results were calculated and corrected for the mass of boron on blank wipes. The average recovery efficiency for the cotton/polyester cloth was 88% while 77% recovery was obtained from the proprietary moist wipe. However, the latter were more convenient and were used in the study.

Dermal exposure to MWF was measured on between one and four occasions throughout the working day depending on the work schedule and work breaks. This was done by wipe sampling each hand separately. The wipes from the samples collected from each individual throughout the day were combined together for analysis. All samples were transported to the laboratory in clean sealed containers. Field blanks and laboratory blanks were taken for the wipe samples. Approximately four or five of the men who were sampled were also recorded on video for periods of 45 min. This was to enable identification of tasks likely to produce significant amounts of dermal exposure.

Hand wipes were analyzed for boron and results were converted into mass of undiluted MWF. In addition, the mass of in-use (i.e. diluted) MWF was calculated using the MWF sump boron concentration.

In total, there were 196 pairs of measurements of exposures on right and left hands. Data were analyzed on the log scale because of a skewed distribution. A paired t-test on these values showed that the mean difference between right and left hands was not statistically significantly different from zero (P = 0.33). For this paper, we have used the average exposure measurement from both hands.

Further analysis of the data showed that the training intervention carried out by the authors resulted in a reduction in dermal exposure and for this reason, we have restricted the analysis in this paper to the baseline data from each of the sites and the repeat visit data for the three control sites. This resulted in 16 measurements being available for situations where workers were exposed to oil-based fluids and 96 to water-based MWF.

The exposure measurements ranged from 100 to 28 000 mg MWF per hand (front and back) for oil-based fluids and from 100 to 170 000 mg MWF per hand for the water-based fluids, expressed as in-use formulations. The median hand exposures for oil- and water-based MWF were not significantly different, with a median exposure for both of 2600 mg per hand.

From observation of the work practices, the process of individual dermal exposure to MWF was highly variable and dependent on the task but can be summarized as occurring at four main stages of the metalworkers job:

  1. Machine set up often involved handling drill bits and other tools within theater of the cutting machine. This was frequently carried out with items that were coated in MWF from previous use.

  2. Machine operation. Often this was completely automated and consequently there was little direct contact with the MWF. However, in many manual and semi-automated machines the worker moved the MWF nozzle to direct it accurately to the cutting edge. This frequently resulted in short whole hand exposure events.

  3. Work piece removal. On completion of the task the cut item was removed from the tool. This item was coated with MWF and handling was usually done without gloves or without any attempt to remove excess fluid.

  4. Machine/sump maintenance. Inspection of the sump fluid, removal of excess swarf and general machine maintenance gave rise to dermal exposure to MWF.

Clearly, the frequency of the machine set up/operation/removal cycle may have a large influence on the degree of dermal exposure. In some workplaces investigated by Semple et al., a single worker had the responsibility for sump maintenance for all machines and they may therefore have had high exposure to concentrated water-based MWF.

The concentration of water-based MWF in the sump was also measured (by centrifuging a known volume of acidified sample at 2400 r.p.m. for 20 min; the volume of the top layer was then measured and used to calculate the fluid strength as a percentage) as a part of this study and data were available for 93 of the dermal samples. These data ranged from 1 to 15%, with a median of 7%. The results from the exposure measurements and the sump analysis are shown in Fig. 1. The average dermal exposures differed between plants although there is substantial overlap in the data from all of the plants. There is some indication of an association between the percentage of MWF in the mix and the exposure, although the regression only explained ∼11% of the variation in the data (linear regression on the log scales, P < 0.001). This correlation is unsurprising since the percent MWF is used in the calculation of the mass of MWF per hand.

Fig. 1.

Scatter plot of the mass of MWF per hand and the percentage of MWF in the machine sump (water-based fluids).

Fig. 1.

Scatter plot of the mass of MWF per hand and the percentage of MWF in the machine sump (water-based fluids).

Study 2 (Roff et al.)—summary of findings

The study by Roff et al. describes measurements made as part of the RISKOFDERM program. This included dermal exposure measurements obtained from 25 subjects at three different UK engineering works. Unfortunately, the authors were only able to acquire seven measurements of hand exposure, although there were 31 measurements of whole-body potential dermal exposure made using Tyvek oversuits. The hand samples were obtained using cotton sampling gloves worn under protective gloves. The stated reason for the low number of hand samples was because most workers would not wear the gloves, as they were worried that the cotton gloves could become entangled in the machinery. All of the samples were obtained over a short time period, i.e. 18 to 90 min.

Results from the Roff et al. study were reported as in-use formulation, with median loading rate of 3.2 mg cm−2 h−1 with the range from 1.4 to 5.4 mg cm−2 h−1. The geometric mean and geometric standard deviation (GSD) were 2.9 mg cm−2 h−1 and 1.67, respectively. Unfortunately, the authors did not report the total duration of the work and so we have estimated exposure lasted for 6 h (based on our experience of this type of work). The area of both hands was given as 820 cm2. On this basis, the median exposure would have been 16 000 mg in-use MWF on hands with the estimated 90th percentile (based on the reported geometric mean and geometric standard deviation) being 28 000 mg MWF (Note: Roff et al. used 820 cm2 as the area of the hands but for consistency with the risk assessment procedures, we have used 840 cm2 in these calculations.).

The authors realized that these were very high exposures, particularly since the sampling gloves were worn under protective gloves. They say, ‘in retrospect, we should have adopted a hand washing method for MWFs when gloves were refused’. They judged that the sampling media might have had a much greater capacity to retain the MWF in relation to the capacity of the skin, which may have invalidated the results.

Study 3 (van Wendel de Joode et al.)—summary of findings

This study compared two objective measurement methods (absorbent pads worn on the hands, forearms, and neck and a fluorescent tracer method) with a new semi-quantitative predictive model known as DeRmal Exposure Assessment Model (van Wendel de Joode et al., 2005). The study was carried out in four metal machining departments in a Dutch truck manufacturing plant. A total of 51 sets of samples were obtained from 36 workers. The residue collected on the pads were analyzed for ethylene glycol content and the in-use mass per unit area of MWF was calculated. Images of hands, forearms, and neck were collected under UV light and analyzed to assess the mass of tracer per unit area and hence the mass of MWF per unit skin area. The data from both approaches were used to estimate the total mass of MWF on hands and forearms by multiplying by the appropriate area of skin exposed. We assume that the pad samples were collected over a full shift and the tracer measurements were made at the end of the shift, but the authors do not explicitly state this.

The data from the Dutch study was presented as geometric mean and GSD. The geometric mean exposure to in-use MWF measured using the pads was 3700 mg (GSD = 2.5) and using the fluorescent tracer 1400 mg (GSD = 5.5). The corresponding estimated 90th percentiles were both 12 000 mg in-use MWF.

These authors also recognize that the pad sampler method may have overestimated dermal exposure because ‘the sampling material generally tends to capture and retain more contamination than is attached to the skin’.

MCCP exposure assessment

Table 1 summarizes the dermal MWF data from these three sources. Note: ‘typical’ exposure assessment represents the median or geometric mean exposure.

Table 1.

Estimated exposure of hands or hands and forearms to in-use MWF formulation

Study N MWF (mg)
 
Comments 
Typical 90th percentile 
Semple et al. 112 5200 36 000 Water and oil based. No protective gloves. Estimates are for both hands 
Roff et al. 16 000 28 000 Sampling gloves under protective gloves. Both hands 
van Wendel de Joode et al. 51 3700 12 000 Hands and forearms by pads, but protective gloves worn 
 38 1400 12 000 As above, but measured with fluorescent tracer 
Study N MWF (mg)
 
Comments 
Typical 90th percentile 
Semple et al. 112 5200 36 000 Water and oil based. No protective gloves. Estimates are for both hands 
Roff et al. 16 000 28 000 Sampling gloves under protective gloves. Both hands 
van Wendel de Joode et al. 51 3700 12 000 Hands and forearms by pads, but protective gloves worn 
 38 1400 12 000 As above, but measured with fluorescent tracer 

Comparison of the data in Table 1 is not straightforward because of the differences in sampling methodology. It is likely that the interception samplers, i.e. pads or absorbent gloves will have retained much higher levels of MWF than the skin. For this reason, they are likely to overestimate the actual exposure received by the workers and more likely represent an assessment of the total mass of MWF that contacted the skin. It might be argued that the fluorescent tracer would provide a more realistic assessment and one that is comparable to removal techniques. However, this depends on the effectiveness of retention of the tracer on the skin and the efficiency of removal of the MWF during the wiping process. In addition, the removal techniques provide a measure of the contamination retained on the skin at the time of wiping and will not take account of any absorption through the skin or losses due to evaporation or transfer to other surfaces. Neither wiping nor the fluorescent tracer method provides a clear advantage over the other.

The data from Roff et al. and van Wendel de Joode et al. are both from situations where protective gloves were worn and so the measurements represent actual exposure but with the protective effect of the gloves included. Experience and the observations made during the reviewed studies suggest that gloves are not commonly worn in this type of situation and they may not be consistently worn throughout the work shift. van Wendel de Joode et al. found that gloves provided some protection; although since most workers in their study wore gloves, the reliability of their assessment of glove effectiveness is unclear. For the pad samples, the exposure was four times higher when gloves were not worn (seven samples where gloves were not worn) and 18 times higher for the fluorescent tracer data (only three samples where gloves were not worn). The Roff et al. data were on average higher than the other reviewed studies, which may reflect the fact that we assumed a work duration that was perhaps longer than was actually the case. The data from the van Wendel de Joode et al. study with absorbent pads were on average slightly lower than that from Semple et al. and their fluorescent tracer data was about a quarter of the Semple et al. data, which would be reasonably consistent with some protection being afforded by gloves.

The data reported by Semple et al. (2007) provides the largest available set of information about MWF exposure and is the only source that represents exposure data for workers without protective gloves. We have therefore used this source of data to estimate exposure to MCCP from the use of MWF.

For water-based fluids, we first multiplied dermal exposure data relating to the in-use formulation by the associated measured MWF sump concentration to give the mass of neat MWF on the hands. Assuming the typical proportion of MCCP in water-based fluids would be 8%, the typical exposure estimate is 30 mg MCCP on both hands. If the maximum MCCP content was 20%, then the RWC hand exposure (90th percentile) would be 520 mg. Note: these figures are based on the water-based MWF data only.

For oil-based fluids, it is not necessary to adjust the in-use exposure for dilution effects. We assumed there was typically 10% MCCP in the oil-based MWF and at most 70% MCCP content. This gives a typical exposure of 520 mg MCCP on both hands and a RWC estimate (90th percentile exposure with highest proportion MCCP) of 25 000 mg MCCP on the hands.

Semple et al. (2007) only measured exposure on the hands and there are no data for other parts of the body. van Wendel de Joode et al. (2005) measured the exposure on the face and neck using the fluorescent tracer method and found that only 32% of samples had detectable levels compared with 92% of the hand samples. Roff et al. (2004) have more information about potential dermal exposure on the body, i.e. measurements made on the outside of work clothing. Their median potential body exposure rate was 0.12 mg cm−2 h−1, which was ∼4% of the actual exposure to the hands. The experience of Semple et al. with patch samplers worn on the wrist of subjects where they detected very low levels supports the view that the vast majority of MWF dermal exposure occurs on the hands. Given these data and the protective effect of the working clothing in reducing body exposure, we judge that only hand exposure makes an important contribution to dermal exposure to MWF, although this clearly relies to some extent on good working practices minimizing potential exposure elsewhere on the body.

Our estimates of typical and RWC daily dermal exposure are summarized in Table 2.

Table 2.

Typical and RWC estimates of daily dermal MCCP exposure

Product MCCP (mg) (both hands—front and back)
 
Typical RWC 
Water-based MWF 30 520 
Oil-based MWF 520 25 000 
Product MCCP (mg) (both hands—front and back)
 
Typical RWC 
Water-based MWF 30 520 
Oil-based MWF 520 25 000 

DISCUSSION

MCCPs are high molecular weight viscous liquids with very low vapor pressures. It is therefore likely that uptake through the skin following dermal exposure is quite low. In the European risk assessment, it was assumed that ∼1% of the mass of MCCP on the skin will be absorbed and this is substantiated by limited experimental data.

While there are some exposure data for specific component chemicals in MWFs (Henriks-Eckerman et al., 2007), there are no dermal MCCP exposure measurement data available. As a result, the European risk assessor had initially used the Estimation and Assessment of Substance Exposure (EASE) expert system (Tickner et al., 2005) to make an assessment of the likely exposure. The EU risk assessment assumed that the most appropriate EASE scenario was ‘non-dispersive’ use with direct and extensive handling, for which the predicted dermal exposure lies in the range 1–5 mg cm−2 day−1 (840–590 000 mg day−1). Taking into account the fact that the MCCP is present at a maximum concentration of 20%, and for water-mix MWF is diluted for use (say to 7%), the predicted range of dermal exposure for use of water-based MWF would be 0.014–0.07 mg cm−2 day−1 over 840 cm2 (hands only). The predicted worst-case exposure is then 59 mg MCCP per day. For oil-based fluids, there is no dilution of the fluid and the predicted range of MCCP from EASE would be 0.7–3.5 mg cm−2 day−1, assuming 70% MCCP in the fluid. The predicted worst-case exposure in this situation being 2940 mg day−1.

It is well recognized that the EASE model may give assessments of exposure that are generally higher than would be encountered in realistic workplace situations e.g. Creely et al. (2005) and Hughson and Cherrie (2005). In this review, we have identified a suitable surrogate dataset from Semple et al. (2007) that can be used to estimate exposure to MCCP. We believe that this approach is likely to be the most suitable given that directly relevant data are not available.

Review of the published literature has identified two other studies (Roff et al., 2004 and van Wendel de Joode et al., 2005) where dermal exposure to MWF was measured. However, these measurements were carried out with protective gloves being worn by most of the subjects and they are therefore not suitable to assess exposure for typical workplaces, where in our experience gloves are not routinely worn.

There are no measurement methods that provide a clear advantage in measuring dermal exposure to MWF. Interception samplers such as the pads used by Roff et al. and van Wendel de Joode et al. are likely to overestimate exposure and the comparison of data from the three studies, summarized in Table 1, tend to support this view. The fluorescent tracer method provides some advantages in terms of convenience but may still overestimate exposure if the fluorescent compound preferentially binds to the skin, as is often the case. The wipe sample may underestimate exposure because of removal of contaminant prior to sampling from washing or wiping of hands or because of uptake of the contaminant through the skin. However, on balance, we believe that the method used in the Semple et al. study is suitable for the present purpose.

We have employed a number of conservative assumptions in arriving at our estimated exposure levels. Where there were ranges of possible values for the proportion of MCCP in MWF, we have generally taken the higher value to estimate the RWC and an ‘average’ value for the typical exposure. The RWC exposures were all based on the estimated 90th percentile of the data so they represent extreme exposure that it is unlikely to be experienced by a given individual on a daily basis.

On this basis, we identified typical daily exposure to MCCP in the use of MWF of 30 mg for water-based fluids and 520 mg for oil-based fluids. The higher estimate for oil-based products arises from the greater proportion of MCCP typically found in these and because they are used without dilution. The corresponding RWC estimates were 520 and 25,000 mg per day for water- and oil-based products, respectively. Note: these figures are higher than the corresponding EASE estimates (i.e. 59 mg day−1 and 2940 mg day−1).

As we indicated earlier, MCCP is likely to be poorly absorbed via the skin, although there is only limited information available. Scott (1989) describes data for one sample of MCCP tested in an in vitro system using excised human skin where he measured a mean absorption rate of 0.04 μg cm−2 h−1. Assuming the whole of both hands were exposed for 8 h, then this would result in an average MCCP uptake of 0.27 mg day−1. The EU risk assessment report makes reference to an unpublished study of the in vitro absorption of a sample of MCCP through human epidermis (by Johnson, 2005), in which a maximum rate of 0.057 μg cm−2 h−1 was measured following occluded application. The EU report quotes the maximum measured dermal absorption of the radio-labeled C15 marker in this study to be 0.7%. Using the exposure data summarized here and assuming a 1% absorption, as was proposed in the EU risk assessment, would give an average dermal MCCP uptake of 0.3 mg day−1 for water-based fluids.

While it was not our intention to estimate inhalation exposure, it is interesting to compare the estimated uptake through the skin with possible inhalation uptake from MWF use. Simpson et al. (2003) reported MWF inhalation exposure data from 31 different worksites, with 40 personal exposure measurements for inhalable oil-based fluids and 298 for water-based fluids. The geometric mean exposure level for oil-based fluids was 0.67 mg m−3 (GSD 3.26) and for water-based fluids 0.13 mg m−3 (GSD 3.9). Assuming that workers inhaled 12.8 m3 of air per working day, that 50% of the inhaled MCCP was systemically absorbed (SCHER, 2008) and using the assumptions about the proportion of MCCP in MWF used in this paper would give an estimated average uptake of 0.45 mg day−1 for oil-based fluids and 0.05 mg day−1 for water-based fluids. These data suggest that dermal uptake of MCCP is at least as important as inhalation uptake.

In conclusion, we have estimated typical (median) and RWC (90th percentile) exposure for MCCPs in MWF. This process was necessary because there is no relevant measurement data available, which in our experience is increasingly the case when regulatory risk assessments are required. While there is considerable uncertainty in the exposure assessment process used in this case, we have tried to ensure that we have tended to overestimate rather than underestimate exposure. In doing this, we have ended up with RWC estimates that were higher than predicted by the EASE model. It has been previously suggested that the EASE model overestimates dermal exposure (Hughson and Cherrie, 2005), but this may not always be the case particularly when estimating RWC exposures. There is an urgent need to develop a reliable model of dermal exposure that can be used in regulatory risk assessments, particularly in relation to the REACH Regulations. In the meantime, we believe that using analogous data provides the best approach to identify the typical and RWC dermal exposures.

FUNDING

The EuroChlor Chlorinated Paraffin Sector Group.

We are grateful for comments on this work from our colleagues Hilary, Cowie, David Farrar, Graeme Hughson, Christine Northage, Anne Sleeuwenhoek, and Martie van Tongeren.

References

Creely
KS
Tickner
J
Soutar
AJ
, et al.  . 
Evaluation and further development of EASE model 2.0
Ann Occup Hyg
 , 
2005
, vol. 
49
 (pg. 
135
-
45
)
Eisen
EA
Smith
TJ
Kriebel
D
, et al.  . 
Respiratory health of automobile workers and exposures to metal-working fluid aerosols: lung spirometry
Am J Ind Med
 , 
2001
, vol. 
39
 (pg. 
443
-
53
)
Friesen
M
Costello
S
Eisen
E
Quantitative exposure to metalworking fluids and bladder cancer incidence in a cohort of autoworkers
Am J Epidemiol
 , 
2009
, vol. 
169
 (pg. 
1471
-
8
)
Henriks-Eckerman
ML
Suuronen
K
Jolanki
R
, et al.  . 
Determination of occupational exposure to alkanolamines in metal-working fluids
Ann Occup Hyg
 , 
2007
, vol. 
51
 (pg. 
153
-
60
)
Hughson
GW
Cherrie
JW
Comparison of measured dermal dust exposures with predicted exposures given by the EASE expert system
Ann Occup Hyg
 , 
2005
, vol. 
49
 (pg. 
111
-
23
)
Robertson
W
Robertson
AS
Burge
CB
, et al.  . 
Clinical investigation of an outbreak of alveolitis and asthma in a car engine manufacturing plant
Thorax
 , 
2007
, vol. 
62
 (pg. 
981
-
90
)
Roff
M
Bagon
DA
Chambers
H
, et al.  . 
Dermal exposure to electroplating fluids and metalworking fluids in the UK
Ann Occup Hyg
 , 
2004
, vol. 
48
 (pg. 
209
-
17
)
SCHER
Risk assessment report on alkanes, C14-17, chloro MCCP, human health part
 , 
2008
Brussels, Belgium
European Commission
 
Scott
RC
In vitro absorption of some chlorinated paraffins through human skin
Arch Toxicol
 , 
1989
, vol. 
63
 (pg. 
425
-
6
)
Semple
S
Graham
M
Cowie
H
, et al.  . 
The causative factors of dermatitis among workers exposed to metalworking fluids.
 , 
2007
Sudbury
HSE Books
 
Available at: http://www.hse.gov.uk/research/rrpdf/rr577.pdf. Accessed 20 November 2009
Simpson
AT
Stear
M
Groves
JA
, et al.  . 
Occupational exposure to metalworking fluid mist and sump fluid contaminants
Ann Occup Hyg
 , 
2003
, vol. 
47
 (pg. 
17
-
30
)
Sprince
NL
Palmer
JA
Popendorf
W
, et al.  . 
Dermatitis among automobile production machine operators exposed to metal-working fluids
Am J Ind Med
 , 
1996
, vol. 
30
 (pg. 
421
-
9
)
Tickner
J
Friar
J
Creely
KS
, et al.  . 
The development of the EASE model
Ann Occup Hyg
 , 
2005
, vol. 
49
 (pg. 
103
-
10
)
van Wendel de Joode
B
Bierman
EP
Brouwer
DH
, et al.  . 
An assessment of dermal exposure to semi-synthetic metal working fluids by different methods to group workers for an epidemiological study on dermatitis
Occup Environ Med
 , 
2005
, vol. 
62
 (pg. 
633
-
41
)
Vermeulen
R
Heideman
J
Bos
RP
, et al.  . 
Identification of dermal exposure pathways in the rubber manufacturing industry
Ann Occup Hyg
 , 
2000
, vol. 
44
 (pg. 
533
-
41
)