Search tips
Search criteria 


Logo of oenvmedOccupational and Environmental MedicineVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Occup Environ Med. 2007 December; 64(12): e11.
PMCID: PMC2095346

Exposure assessment 1

048 Chemical exposure histories recorded in 18 000 Porton Down veterans

T. J. Keegan1, C. Brooks1, S. Walker1, T. Langdon1, P. Doyle2, N. E. S. Maconochie2, T. Fletcher2, M. J. Nieuwenhuijsen3, L. M. Carpenter1, K. M. Venables1, T. J. Keegan1. 1University of Oxford; 2London School of Hygiene and Tropical Medicine; 3Imperial College, London

ObjectivesResearch into the effects of chemical warfare agents on military capability at Porton Down started in 1916. We assembled exposure histories for a cohort of servicemen who took part in tests between 1941 and 1989. The methods used to identify, assign and categorised chemical exposures to cohort members will be described.

MethodsData on tests were abstracted from the historical archives by a team of four research staff and entered into a database. Each test was linked to a cohort member (veteran) using the individual's name and date(s) they were present for test(s). Where possible, chemicals were assigned a Chemical Abstract Service (CAS) number. A classification system was developed reflecting the intended use with reference to the Chemical Weapons Convention and other published sources.

ResultsTest data were linked to 18 184 identifiable veterans. The annual number of cohort members tested varied over time; most (75%) were first tested in the 1940s, 1950s or 1960s. Overall, veterans took part in approximately 200 000 tests, of which 150 000 involved chemicals and 50 000 did not (eg, equipment trials). Of the 1800 unique chemicals recorded, a CAS number was assigned to 20% and it was possible to classify 95% according to intended use. Vesicants were the most frequently‐tested group of chemicals (57 000 tests, including tests for dermal hypersensitivity to sulphur mustard) followed by therapies (27 000 tests) and riot control agents (10 000 tests). Overlap between groups, availability of quantitative “dose” data, and of information on use of respiratory and other protection, will be reported.

ConclusionServicemen included in a cohort study of Porton Down veterans were involved in tests of over 1800 chemicals between 1941 and 1989. Classifying chemicals using a system based on the intended use, the most commonly tested substances were vesicants, therapies, and riot control agents.

Key wordsexposure assessment; cohort study; chemical warfare

049 Mercury exposure assessment for workers in fluorescent‐tube recycling plants

C. M. Hsiech1, H. Y. Yang2, T. S. Shih1, Y. C. Lin3. 1Institute of Occupational Safety and Health; 2Institute of Environmental Health, National Taiwan University; 3Taiwan Toko University

ObjectivesAbout 90 million fluorescent lamps are disposed of each year in Taiwan. Since each of those fluorescent tubes may contain as much as 10 mg of mercury, the hazardous waste thus generated causes a serious impact on the environment. The Taiwan Environmental Protection Agency began the recycling of fluorescent tubes in 2002 and four factories are authorised to process the waste lamps. The recycling of fluorescent lamps is a newly‐established industry and there are no worker exposure assessment data available in Taiwan.

MethodsIn our study, the assessment of worker exposure to mercury was conducted by means of workplace ambient air monitoring, biological monitoring and collection of questionnaire data.

ResultsThe fluorescent tubes were broken down and the materials were recovered in an enclosed processing unit where a negative pressure was maintained. For stationary samples collected for an 8 h time weighted average (TWA) measurement outside the unit, the mercury vapour concentrations were between 0.0014 and 0.018 mg/m3 (n = 6). The mercury vapour concentrations were between 0.0014 and 0.40 mg/m3 (n = 34) for samples near the outlets of the processing unit; concentrations inside the unit were found to be between 0.12 and 3.0 mg/m3 (n = 12) during operation, and between 0.016 and 0.70 mg/m3 (n = 11) during clean‐up. Mercury vapour concentrations of personal samples for the lamp‐loading workers were between 0.0014 and 0.18 mg/m3 (n = 11), and were affected by the frequency with which workers entered the unit. Airborne fluorescent dust was collected in PdCl2‐coated MCE filters and results of measurement revealed that a substantial amount of mercury vapour was absorbed in the dust. Mercury exposure may be underestimated if mercury vapour alone is analysed and mercury absorbed in the dust is neglected. Although all recycling plant workers wear respirators, improper use or maintenance might result in exposure to mercury.

ConclusionAccording to the results of the questionnaire, biological monitoring and ambient air monitoring, we concluded that whether or not respirators were correctly used was the main factor affecting mercury exposure among workers.

Key wordsfluorescent lamp; mercury; exposure assessment

050 Estimation of parameters in logistic regression models with multiplicative measurement error: individual‐ and group‐based exposure assessments

H. M. Kim, I. Burstyn. University of Alberta

ObjectivesLognormal distribution of measured exposure levels suggests that measurement error is multiplicative. The group‐based strategy in which each subject is assigned the median value of exposure of their group can be used to assess exposures in such situations. In contrast, an individual‐based strategy can be applied when we have an exposure measurement for each subject. Regression calibration is an effective method of correcting estimates of parameter in regression models for measurement error when its assumptions are satisfied. We aimed to evaluate bias in logistic models when group means are fixed with group‐ and individual‐based exposure assessment.

MethodsWe conducted computer simulation studies to compare naive and regression calibration estimates in individual‐based strategy with estimates obtained using group medians in group‐based strategy. We varied between‐ and within‐subject variances, and magnitude of group means, but held between‐group differences constant and true value of risk parameter (0.3). Measurements of 100 subjects per group were used to estimate group medians.

ResultsBias in risk estimate depended on both the between‐ and within‐subject variability for individual‐based assessment. For example, for “small” group means attenuation increased from 0.05 to 0.24 as within‐subject standard deviation (SD) increased from 0.3 to 1 (between‐subject SD 0.3); attenuation decreased from 0.05 to 0.03 as between‐subject SD increased from 0.3 to 1 (within‐subject SD 0.3). Only the between‐subject variability influenced bias in group‐based strategy. For example, for “small” group means the overestimate increased from 0.1 to 0.7 as between‐subject SD increased from 0.3 to 1. Regression calibration performed well in adjusting risk estimates obtained in individual‐based analyses. For fixed values of between‐ and within‐ subject SD (0.3), when group means were “small”, the regression calibration overestimated the true risk parameter by 0.04; for “large” group means, regression calibration estimate was attenuated by 0.11. When group means were “large” and between‐subject SD was at least moderate ([gt-or-equal, slanted]0.7), all approaches resulted in considerable attenuation.

ConclusionBoth the individual‐ and group‐based strategies can lead to biased risk estimates when exposures are grouped and measurement error is multiplicative. Regression calibration estimates usually have less bias, but their performance depends on the absolute value of true exposure in each group.

Key wordslognormal exposure; bias; log‐linear model

051 Job exposure matrix of silica for women in Shanghai

A. T. Huang, W. H. Chow, J. Coble. Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, NIH, DHHS

ObjectiveThe objective of this study was to develop a job exposure matrix (JEM) to estimate occupational levels of silica exposure for a study of women in Shanghai, China.

MethodsThe Shanghai Women's Health Study (SWHS) enrolled 74,942 adult Chinese women in Shanghai during 1996‐2000. Nearly 100% of these women were employed outside the home. A silica JEM was merged with job histories by both occupation and industry titles as well as silica‐dust measurement database. A database of silica‐dust measurements sampled during 1954–2000 was used to estimate exposure levels by JEM intensity rating. Regression analysis was used to evaluate trends in silica‐dust level over time and to model silica‐dust levels by time period. In order to increase the specificity of JEM, the exposure criteria based on JEM ratings and the number of measurements categorized the study subjects into three groups: the most restrictive high exposure group, uncertain exposure group, and reference group. The most restrictive high exposure group was identified as those having high probability of exposure from both occupations and industries. The reference group was defined as women who had never been employed in occupation or industry with potential silica‐dust exposure. The remaining study subjects were categorized as the uncertain exposure group. This job quantitative exposure matrix was developed to estimate lifetime duration of exposure, lifetime average exposure, and lifetime cumulative exposure for each subject.

ResultsThe results of regression analysis show that silica‐dust levels were reduced substantially from 1954 to 2000. The most restrictive high exposure group (0.7% of the total study subjects) was exposed to an average exposure level of 17.8 mg/m3. The uncertain exposure group (29.3% of the total study subjects) was exposed to an average exposure level of 10.6 mg/m3. Approximate 91% of the studied women in the both of the most restrictive high exposure group and the uncertain exposure group were exposed to silica‐dust higher than 5 mg/m3, OSHA's permissible exposure limit (PEL) for respirable dust containing no crystalline silica.

ConclusionsThe findings suggest that some SWHS participants might have experienced substantial occupational silica exposure in the past. The monitoring data can be used to refine JEM estimates of silica exposure, which can be assessed in relation to respiratory disease risk.

Key wordsjob exposure matrix; silica; exposure

052 Effect of PM exposure on coagulation indexes among steel workers in Italy

M. Bonzini1, A. Baccarelli2, L. Tarantini3, G. Rizzo3, B. Marinelli3, P. A. Bertazzi3, A. Tripodi4, A. Artoni4, P. M. Mannucci4, P. Apostoli5. 1Department of Occupational and Environmental Health, Foundation IRCCS Maggiore Hospital, MaRE, Milan, Italy; 2Exposure, Epidemiology, and Risk Program, Harvard School of Public Health, Boston, Massachusetts; 3Department of Environmental and Occupational Health, University of Milan, Italy; 4Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Department of Internal Medicine, University and IRCCS Maggiore Hospital, Milan, Italy; 5Institute of Occupational Health and Industrial Hygiene, Brescia University, Italy

ObjectivesTo verify if occupational exposure to measured levels of particulate matter (PM) causes alteration in inflammatory and coagulation indexes in a group of healthy workers, in order to better understand the biological mechanisms by which environmental exposure to PM provokes an increase in cardiovascular risk in the general population.

MethodsWe enrolled 63 male workers employed in a steel production plant near Brescia (Italy). Blood drawings in the morning of the first day and of the last day of the working week were obtained for each subject. Detailed information about smoking habits, non‐occupational exposure to PM, working and health history were collected by questionnaire and from clinical records. Personal exposure to PM10 and PM1 was calculated considering results of environmental monitoring in the workplace and the specific tasks of each subject. Multiple regression was used to analyse the relationship between PM exposure and coagulatory/inflammatory parameters, after adjustment for non‐occupational confounders (age, BMI, education, smoking and NSAIDs consumption).

ResultsWorkers enrolled were exposed to PM levels from 2 to 20 times higher than the general population (mean PM10 233 μg/m3, range 73–1220). All inflammatory/coagulatory indexes (CRP, platelet function analyser (PFA100), endogenous thrombin potential (ETP), PT, aPTT, t‐PA, PAI, D‐dimer) results were normal. We observed a relationship between PM exposure during the working week and increase in closure time (PFA100) (respectively 80, 90 and 94 s in tertiles of PM1 exposure, p value from regression = 0.02). No dose–response relation was observed for any of the other parameters considered. When differences in haematic parameters were analysed (beginning–end of working week), an increase in PFA100 closure time and a slight non‐significant decrease in ETP, t‐PA and PAI were found to be associated with increasing exposure.

ConclusionOur study conducted among workers exposed to high levels of PM did not find a clear relationship between personal exposure and pro‐coagulatory tendency. We observed an unexpected increase in PFA100 closure time (index of platelet function) mainly related to PM1 exposure. We plan to enlarge the study population and to consider more haematic indexes (adhesion molecules and platelet microparticle analyses are on going) to better understand how PM exposure affects coagulation processes.

Key wordsoccupation; fine particles; coagulation

053 An assessment method for evaluating parental occupational exposure to specific agents: the United Kingdom Childhood Cancer Study (UKCCS)

O. Y. Raji1, M. van Tongeren2, R. G. Feltbower1, P. A. McKinney1. 1Paediatric Epidemiology, University of Leeds; 2Institute of Occupational Medicine

ObjectivesTo design and evaluate a retrospective assessment procedure to characterise reported parental occupational exposures.

MethodsThe UKCCS, a national case‐control study of childhood cancer, recorded detailed occupational histories of both parents with closed questions on exposure to eight specific agents (solvents, paints/thinners, dyes/pigments, petroleum products, lead compounds, fertilisers/pesticides, radiation/radioactive materials, and wood/other dust) for each of their previous jobs. The review method took all positive responses to the specific questions and examined information on job titles, industrial activities, task descriptions, and usage of the materials. These exposure determinants were scrutinised methodically to provide subjective semi‐quantitative opinions on exposure probability, level of exposure, frequency of contact, and degree of protection. The expressed opinions on the four measures were combined systematically to categorise parental exposure status as “exposed”, “partially exposed” or “unexposed”.

ResultsA total of 22 319 exposure instances were individually evaluated (mothers = 4408, fathers = 17 761, other relatives = 150). Exposure to solvents was most commonly reported overall (25.8%), and by fathers (25.0%) and mothers (29.4%) separately, while fertilisers/pesticides exposure was the lowest reported agent overall (5.1%). The review procedure suggested that 39.3% of the exposure instances could be categorised as “exposed”, 55.4% as partially exposed and the remaining 5.3% as unexposed. The false positive exposure was highest for mothers (15.1%) and exposure to radiation/radioactive materials (34.4%).

ConclusionThe exposure review procedure demonstrates that systematic examination of detailed occupational data increases the potential for distinguishing between individuals who were likely to be exposed to specific agents from those reporting trivial exposures. The results show that the number of false positive exposures in population‐based case‐control studies will reduce if exposure occurrences directly reported are combined with other detailed exposure and occupational information in various work activities to ascertain parental exposure status.

Key wordsparental occupational exposure; retrospective assessment; childhood cancer

054 Accuracy of toxicological information on material safety data sheets for toluene diisocyanate‐containing products used in workplaces in Canada

L. ElBilali1, P. Demers2, M. Nicholas3. 1National Office of Workplace Hazardous Materials Information System; 2School of Occupational and Environmental Hygiene, University of British Columbia; 3National Office of Workplace Hazardous Materials Information System, Health Canada, Ottawa, ON

ObjectivesMaterial safety data sheets (MSDSs) are one of the primary means of hazard communication that allow workers to safely use, handle and dispose of hazardous chemicals. MSDSs in Canada, solely a Workplace Hazardous Materials Information System (WHMIS) requirement, are established under the Hazardous Products Act (HPA) and the Controlled Products Regulations (CPR) and administered at the national level by Health Canada's WHMIS program conterminously with all provincial/territorial governments' and Human Resources and Skills Development's occupational health and safety regulatory agencies administering WHMIS employer requirements within their respective jurisdictions. Exposure to toluene diisocyanate (TDI, CAS No. 584‐84‐9) is a recognised cause of occupational asthma. TDI, a respiratory sensitiser and irritant is also classified as a carcinogen under WHMIS in Canada by virtue of its IARC designation. Here, the accuracy of the toxicological information, disclosed on MSDSs for TDI containing products used in Canada, is reviewed.

MethodsMSDSs (n = 203, 174 from Canada, 26 from the US, and 3 others) of TDI‐containing products, available in the Canadian Centre for Occupational Health and Safety database, were used. An audit form was developed to review the accuracy of the MSDS hazard information based on the WHMIS Guidelines for the Disclosure of Toxicological Information on a MSDS with respect to MSDS sub‐items: 1(3) ‐ LD50 (species and route); 1(4) ‐ LC50 (species and duration of exposure); 7(1) ‐ routes of entry, including skin contact, skin absorption, eye contact, inhalation and ingestion; 7(2) ‐ effects of acute exposure to product; 7(3) ‐ effects of chronic exposure to product; 7(4) ‐ exposure limits; 7(5) ‐ irritancy of product; 7(6) ‐ sensitisation (respiratory or skin) to product; and 7(7) ‐ carcinogenicity.

ResultsMain results showed the following discrepancies (% indicated in parentheses): routes of exposure (13%); exposure limits, e.g. TLV (25%); carcinogenicity (40%); asthma (26%); acute/chronic exposure effects ((91%‐wheeze), (60%‐delayed effects) and (57%‐aggravation of pre‐existing respiratory conditions)).

ConclusionEnsuring the safe use/handling of TDI in Canadian workplaces requires enforcement of MSDS compliance with WHMIS (ie, HPA/CPR) requirements.


Articles from Occupational and Environmental Medicine are provided here courtesy of BMJ Publishing Group