Skip Navigation



Occupational Medicine Advance Access published online on April 16, 2007
Occupational Medicine, doi:10.1093/occmed/kql153
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
57/4/238    most recent
kql153v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Jones, S.
Right arrow Articles by Walker, P
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jones, S.
Right arrow Articles by Walker, P
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Lung cancer mortality at a UK tin smelter

SR Jones1, P Atkin1, C Holroyd2, E Lutman1, J Vives i Batlle1, R Wakeford3 and P Walker1

1 Westlakes Research Institute, Westlakes Science and Technology Park, Moor Row, Cumbria CA24 3LN, UK
2 Rio Tinto plc, 6 St James' Square, London SW1Y 4LD, UK
3 Dalton Nuclear Institute, University of Manchester, M60 1QD, UK

Correspondence to: Steve Jones, Westlakes Scientific Consulting, Westlakes Science and Technology Park, Moor Row, Cumbria CA24 3LN, UK. Tel: +44 1946 514003; fax: +44 1946 514091; e-mail: steve.jones{at}westlakes.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
 References
 
Background An earlier study of mortality among male former employees at a tin smelter in Humberside, UK, had identified excess mortality from lung cancer, which appeared to be associated with occupational exposure.

Aims The aim of the present study was to investigate the relationship between lung cancer mortality and quantitative measures of exposure.

Methods Using available records of occupational hygiene measurements, we established exposure matrices for arsenic, cadmium, lead, antimony and polonium-210 (210Po), covering the main process areas of the smelter. We established work histories from personnel record cards for the previously defined cohort of 1462 male employees. Three different methods of extrapolation were used to assess exposures prior to 1972, when no measurement results were available. Lung cancer mortality was examined in relation to cumulative inhalation exposure by Poisson regression analysis.

Results No significant associations could be found between lung cancer mortality and simple cumulative exposure to any of the substances studied. When cumulative exposures were weighted according to time since exposure and attained age, significant associations were found between lung cancer mortality and exposures to arsenic, lead and antimony.

Conclusions The excess of lung cancer mortality in the cohort can most plausibly be explained if arsenic is the principal occupational carcinogen (for which the excess relative risk diminishes with time since exposure and attained age) and if there is a contribution to excess mortality from an enhanced prevalence of smoking within the cohort. The implications of the dose–response for arsenic exposure for risk estimation merit further consideration.

Keywords      Arsenic; cohort study; lung cancer; mortality; tin smelter


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Conflicts of interest
 References
 
An earlier paper [1] reports a significant excess in mortality from lung cancer [62 deaths, standardized mortality ratio (SMR) 161, 95% confidence interval (CI) 124–206, P < 0.001] in a cohort of 1462 male employees at a former tin smelter located in Humberside, UK. No data on smoking habits were available for the cohort. Nonetheless, a substantial proportion of the excess could best be attributed to exposure to one or more occupational carcinogens, for which the effect on mortality diminishes with time since exposure. No significant excesses in mortality for any other cause of death were found.

Workers at the smelter were potentially exposed to a variety of substances including tin, lead, antimony, arsenic, cadmium, sulphur dioxide, natural series radionuclides and combustion products. The radionuclide of most interest is polonium-210 (210Po).

For both arsenic and ionizing radiation, causative relationships with excess risk of lung cancer are well established [2–5], while for cadmium the evidence published is conflicting [6–12]. Evidence of a role for lead or antimony inhalation in lung carcinogenesis is weak; where associations have been observed confounding with arsenic is likely [13–16].

The objective of this study was to investigate the relationships between excess lung cancer mortality at the smelter and inhalation exposures to lead, antimony, arsenic, cadmium and radioactivity, with the aim of identifying the cause or causes of the excess.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Conflicts of interest
 References
 
The Capper Pass smelter took process residues from other smelters, together with some raw ores, as feedstock. The main process circuit comprised raw materials handling, feed mixing and drying, sintering and reduction in a blast furnace to produce an impure alloy of tin and lead. The impure alloy was dry refined prior to electrolytic separation which yielded pure tin as a cathode deposit and lead, together with other trace metals, as an anode deposit. The anode deposits were further dry refined and electrorefined to yield pure lead, bismuth/lead alloy and antimony/lead alloy.

Air sampling was carried out both by works staff and the UK regulatory agencies. Area (or static) samples were taken weekly over a 15-min period at set locations. Personal samples were taken on a campaign basis from selected areas, most being taken over a full working shift; the few samples taken over a short period to assess exposures on specific tasks were omitted from the analysis. The available records do not specify the size fractions of the aerosol sampled. The recorded measurements cover the period 1972–91.

Over 20 000 individual sampling results are available. The most comprehensive results are for lead in air, as the smelter was regarded as a ‘lead process’ under UK regulations. Gaps in the data for other substances were filled by making use of ratios to lead derived from available measurements in individual process areas.

‘Area’ or ‘static’ air samples may not be representative of the air breathed by workers and are likely to underestimate true personal exposures [17–20]. Ratios between personal and area sample results from personal sampling campaigns have been used to estimate exposures in terms of the annual mean of personal sample results as measured over a working shift.

Sample data were distributed approximately log normally. The statistic adopted as an annually averaged estimate of exposure was the minimum variance-unbiased estimate of the arithmetic mean [21]. In many cases, data sets contained a proportion of values reported at the limit of detection; a maximum likelihood method [22] was used to obtain the best estimate of the mean.

Annual average exposures in the principal process areas over the period 1972–91 have been summarized as a matrix. The data were insufficient to provide further resolution at the job-within-process-area level. Derivation of the matrix has been reviewed by a panel of former technical managers from the smelter. A more detailed description of the matrix can be found elsewhere [23].

Personnel record cards permitted definition of work histories for each cohort member in terms of the proportion of time spent annually in each process area.

Temporal extrapolation is necessary in estimating exposures, as some employment histories extend back to 1937. Major new process plants and process ventilation systems were introduced from 1967 onwards. This may well have improved working conditions, but production also increased progressively from 1937 onwards [24]. Only one document gives results for air contamination prior to the 1970s; a letter in 1960 from the Factory Inspectorate cited measurements of lead that were substantially higher than those recorded during the 1970s.

Recognizing the uncertainty in early air contamination levels, we have used three scenarios for back-extrapolation in the regression analyses as follows:

Scenario A: Constant back-extrapolation in each process area, as the mean of the levels in the three earliest years for which data were available.

Scenario B: Back-extrapolation in each process area on a linear increasing trend from the baseline value above, to values 2-fold higher in the early 1940s, based on a weak trend seen in per-caput average exposure levels over the period 1972–91.

Scenario C: Back-extrapolation in each process area from the baseline value above to values 2-fold higher in 1960, subsequently, declining linearly to values one-half of the baseline in 1937. In this scenario, air contamination levels initially increase as a consequence of increasing production and ageing process plant, before declining as a consequence of regulatory pressure and capital investment during the 1960s and 1970s.

Individual exposure histories could then be obtained by cross-reference of the exposure matrix and the work histories.

The 35 942 person-years at risk for the 1462 subjects were accumulated into strata defined by age, calendar year and categories of cumulative exposure. Accumulation of person-years commenced for each individual either 12 months after their start of employment date or 1 November 1967 if that date was later. Cumulative exposure categories were defined as quintiles of the exposure distribution experienced by the lung cancer cases, and the point value of exposure assigned to each band was determined as the person-time weighted mean; this is expected to minimize the potential problems associated with exposure misclassification [25]. We calculated expected numbers of deaths from lung cancer within each stratum using the male mortality rates for England and Wales [26].

We performed Poisson regressions on mortality data stratified by cumulative exposure to lead, arsenic, cadmium and 210Po using the EGRET software package [27].

Most studies of this type accumulate person-years into strata defined by cumulative exposure, implicitly assuming that excess relative risk is proportional to the cumulative exposure throughout follow-up. In reality, dose–response relationships are likely to be subtler than this. The earlier study of this cohort [1] found evidence of a diminution of lung cancer risk with time since exposure. Such an effect is well established for tobacco smoke [28–30] and ionizing radiation and has also been suggested for arsenic [31].

Effects of this nature also emerge naturally from process models of carcinogenesis [32–35].

If excess relative risk does vary with time since exposure, regression analyses that assume simple proportionality with cumulative exposure during the follow-up period may fail to identify associations that actually exist. If the cohort were sufficiently large, the dependence of excess relative risk on factors such as intensity of exposure, time since exposure and attained age could be explicitly evaluated. However, the 62 lung cancer deaths in the Capper Pass cohort are insufficient to permit this.

The US Committee on the Biological Effects of Ionising Radiations (BEIR) studied >500 lung cancer deaths in cohorts of uranium miners exposed to radon daughters, and in its sixth report (Biological Effects of Ionising Radiations VI) developed a risk model in which weightings are used to modify cumulative exposure as an explanatory variable for excess relative risk, as [36]:

Formula
where ERR is the excess relative risk, Ei is the exposure during the ith exposure episode, {theta}(ti) is a weighting factor depending on ti (the time since the exposure episode occurred) and {phi}age is a weighting factor depending on attained age. Weights are specified as step functions, reflecting the categorization of person-years at risk.

If we make the assumption that the weighting factors are largely a reflection of the carcinogenic process for lung, rather than the properties of the specific carcinogen, then one may expect excess relative risk to vary in this way for many carcinogenic agents that can, like radon daughters, only exert a modifying effect on the carcinogenic process during the period of exposure. Arsenic (because of its very rapid clearance from the body after intake) is such a case. We have therefore investigated the effects of both the stepwise-specified BEIR VI weighting factor and the smoothed weights shown in Figure 1, on our regression analysis.


Figure 1
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. BEIR VI weighting factors for time since exposure (a) and attained age (b).

 
As an alternative, we have also examined the effects on the regression analysis of assuming that the excess relative risk of lung cancer is only influenced by specific ‘windows’ of time during the prior exposure history.

As no data are available for the smoking habits of the cohort, all the analyses are subject to the implicit assumption that smoking prevalence is not strongly correlated with exposures.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Conflicts of interest
 References
 
Our exposure matrix contains estimated air contamination levels in >20 distinct process areas, together with levels in non-process areas estimated from results of samplers that were operated outdoors around the site. Table 1 provides summary information for the period 1972–91.


View this table:
[in this window]
[in a new window]

  		Table 1. Averagea levels of air contamination in process plants, 1972–1991

 
Table 2 provides summary information on the distribution of individual cumulative exposures under the three pre-1972 exposure scenarios. Simulations based on statistical variance of exposure data and judgements on other sources of error indicate that uncertainty in individual cumulative exposures may broadly be described by a log-normal distribution around the mean, with 95th percentile limits factor of 3 on either side of the mean.


View this table:
[in this window]
[in a new window]

  		Table 2. Distribution of cumulative exposures incurred over the working lifetime of individual cohort members

 
Table 3 provides results from the regression of lung cancer mortality against both unweighted cumulative exposures to lead, antimony, arsenic, cadmium and 210Po and against cumulative exposures weighted for time since exposure and attained age. The results presented are those obtained with smoothed weighting; those obtained with ‘stepwise’ weighting are very similar.


View this table:
[in this window]
[in a new window]

  		Table 3. Results of regression analyses

 
Table 4 shows the regression results for arsenic in more detail; similar data for the other substances may be found as Supplementary data at Occupational Medicine Online.


View this table:
[in this window]
[in a new window]

  		Table 4. Regression results: weighted cumulative exposures to arsenic

 
Significant trends in lung cancer mortality with increasing weighted cumulative exposure to lead, antimony and arsenic are found for all three scenarios of pre-1972 exposure; no significant trends are found in any of the other regressions. For arsenic, antimony and lead, introduction of the weighting increases the gradient, reduces the intercept and increases the log likelihood of the model fit. With the exception of lead in Scenarios B and C, the regressions show projected observed to expected ratios greater than unity at zero exposure.

An alternative analysis in which trends were sought with cumulative arsenic exposure accumulated in 12 defined time windows produced significant trends in only a few instances and no material improvement in regression statistics over those presented in Table 3. The projected observed to expected ratio at zero exposure in these analyses ranged from 1.2 to 1.6.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Conflicts of interest
 References
 
Tables 1 and 2 show modest levels of exposure in comparison to those cited in other studies showing excesses of lung cancer. The studies of copper smelting cohorts [2–4] cite As concentrations exceeding 10 mg m–3 and upper strata of cumulative exposures in excess of 50 mg year m–3. In this study, maximum concentrations were 1.4 mg m–3 and 97.5th percentiles of cumulative exposure 4 mg year m–3. Studies of the cadmium recovery plant cohort [7–9] cite Cd concentrations up to 1.5 mg m–3 and cumulative exposures in excess of 10 mg year m–3, compared with maximum concentrations of 0.8 mg m–3 and 97.5th percentiles of cumulative exposure at 1.5 mg year m–3 for this study. The 97.5th percentile of cumulative exposures to 210Po in this study, at 3 Bq year m–3, implies a radiation dose to lung of 0.2 Sv [37]. Cohorts of uranium miners and nuclear process workers showing substantial excesses in lung cancer mortality have experienced radiation doses to lung in excess of 100 Sv [5].

From Table 3, it is apparent that there are no significant associations between mortality from lung cancer and unweighted cumulative exposures to lead, antimony, arsenic, cadmium or 210Po. In view of the above comparisons, this may not seem surprising.

Comparable regression results for unweighted cumulative exposure to arsenic derived from the published results for copper smelter workers are presented in Table 5.


View this table:
[in this window]
[in a new window]

  		Table 5. Results of comparable regression analyses on arsenic exposure and lung cancer mortality data from copper smelter studies

 
These three cohorts show projected observed to expected ratios at zero exposure that are substantially greater than unity, together with gradients that are statistically compatible with our findings, but which would permit little, if any, of the excess mortality observed in our cohort to be attributed to exposure.

Table 3 shows significant associations between lung cancer mortality and weighted cumulative exposures to lead, antimony and arsenic, with substantial improvements to regression statistics compared with the unweighted analysis, for all three pre-1972 exposure scenarios.

As : Pb ratios in most of the process areas fall within the range of 0.1–0.5; Sb : Pb ratios are ~0.2 with little variance between areas. Thus, there is a substantial degree of correlation between exposures to these three substances. The association between lead or antimony exposure and lung cancer mortality may therefore be attributed to confounding by an underlying causative relationship with exposure to arsenic, or vice versa. Figure 2a shows the ratio of observed to expected mortality in each of the arsenic-exposed strata, together with the estimated values assuming lead to be the real causative factor, according to the lead regression coefficients of Table 3. Figure 2b shows the same data for each of the lead-exposed strata, assuming arsenic to be the real causative factor. Similar results are obtained for antimony. Clearly, the data alone do not permit unambiguous attribution of causality to arsenic exposure, antimony exposure, lead exposure or a combination of the three.


Figure 2
View larger version (14K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Example regression results, exposure Scenario C. (a) Regression against weighted cumulative As exposure, showing predicted mortality on the alternative assumption that exposure to lead is the true causative factor. (b) Regression against weighted cumulative Pb exposure, showing predicted mortality on the alternative assumption that exposure to arsenic is the true causative factor.

 
However, there is strong evidence from other studies for a causative role of arsenic in lung carcinogenesis, while the evidence for lead or antimony is weak. We therefore conclude that the association with weighted cumulative exposures to arsenic reflects an underlying causative relationship, while those for lead and antimony may be attributed to confounding.

In the analysis against weighted cumulative exposures, the highest exposure stratum for 210Po shows a significant elevation in mortality, even though the trend with exposure is not significant (Table 3). However, the high 210Po exposures arise largely in the sinter plants, where arsenic exposures are also high.

The number of deaths in our study (62) is relatively small. Although the power of the study is comparable with that of the Tacoma and Rönnskärversken studies, with 104 and 106 observed deaths, respectively, it is insufficient to permit optimization of the weighting factors for time since exposure and attained age. Nonetheless, some simple sensitivity analyses have suggested that the values used are broadly consistent with the principle of maximum likelihood.

For all three exposure scenarios, the best estimate intercepts for the observed to expected ratio, at zero arsenic exposure, lie between 1.2 and 1.3. These values are consistent with the tendency for smoking prevalence to be higher among manual workers [39] and the range of ratios for lung cancer mortality that have been attributed to enhanced smoking in other groups of manual workers [40–44].

This observation is not inconsistent with the findings of the earlier mortality study [1] that mortality from other smoking-related causes of death is not elevated. The effect of smoking on lung cancer mortality is much greater than that on mortality from other smoking-related causes [45]; if smoking accounted for an observed to expected ratio for lung cancer of 1.2–1.3, the expected ratio for other smoking-related causes would be <1.1. This is well within the CIs of the observed ratios.

If smoking prevalence were correlated strongly with arsenic exposure that might also account for a proportion of the gradient as well as the intercept. However, in this event, smoking would account for a much greater proportional elevation in the mortality of the cohort than has been seen in studies that have controlled for smoking.

Simulations incorporating the assessed uncertainties associated with individual arsenic exposures result in the gradient of the dose–response being reduced by ~10%, while the intercept remains unaltered.

The regression models for weighted cumulative arsenic exposure indicate that in the three exposure scenarios between 11 and 14 lung cancer deaths—rather more than half the observed excess—may be attributed to occupational arsenic exposure.

The question of whether occupational exposure to cadmium, acting either alone or in combination with arsenic, plays a role in lung carcinogenesis has not been satisfactorily resolved [7–12]. As attempts to stratify person-years for the cohort by combinations of low–medium–high arsenic exposure and low–medium–high cadmium exposure were uninformative as a consequence of correlation between exposures, this study offers no evidence for such a role for cadmium alone (Table 3) or for a combined effect of exposure to cadmium and arsenic.

We conclude that a substantial proportion of the excess lung cancer mortality observed in the cohort of workers at the Capper Pass smelter can be attributed to the effects of arsenic exposure, but only if it is assumed that the resulting excess relative risk of lung cancer declines with time since exposure and attained age. The regression results are also consistent with a contribution to the excess from enhanced prevalence of smoking within the cohort. The consistency of these findings across the three exposure scenarios considered lends confidence that they are not an artefact of assumptions made concerning historic exposures.

The possibility that simple cumulative inhalation exposure may not be a good predictor of excess relative risk has important implications for the estimation of the risks resulting from exposure to arsenic. It may be possible to confirm or refute this possibility by re-examining the data from other cohorts of non-ferrous metal smelter workers.

Further, these results raise the hypothesis that similar effects may be seen in cohorts exposed to other lung carcinogens that are cleared rapidly from the body and therefore can only influence processes contributing to carcinogenesis at, or near, the time of exposure.


   Conflicts of interest
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
References
 
All the authors (except R.W. and C.H.) are employed by Westlakes Scientific Consulting Ltd who were funded by Rio Tinto plc to undertake the work reported in this paper. S.R.J. and R.W. are acting as experts, instructed by Rio Tinto plc, in the Capper Pass Claims Review Scheme. C.H. was, during the course of this work, an employee of Rio Tinto plc. Westlakes Scientific Consulting Ltd has also been funded by Rio Tinto to work on other projects both currently and in the past.


   Acknowledgements
 
We are grateful to the late Sir Richard Doll for his wise guidance and constructive criticism throughout the course of this study; to Peter Halsall, John Litten, Mike Fulwell and Mike Perry for their explanations and recollections of the Capper Pass plants and processes; to Tom Sorahan and Robert Lancashire for helpful discussions on the work histories and exposure matrix; to Keith Binks and Michael Gillies for advice on statistical methods; to Les Scott for database management and to Sheila Jones and Marie Dobinson for assistance with ICD coding.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Conflicts of interest
 References
 

  1. Binks K, Doll R, Gillies M, et al. (2005) Mortality experience of male workers at a UK tin smelter. Occup Med (Lond) 55:215–226.[CrossRef][Medline]

  2. Lubin JH, Pottern LM, Stone BJ, Fraumeni JF. (2000) Respiratory cancer in a cohort of copper smelter workers: results from more than 50 years of follow-up. Am J Epidemiol 151:554–565.[Abstract/Free Full Text]

  3. Enterline PE, Henderson VL, Marsh GM. (1987) Exposure to arsenic and respiratory cancer—a reanalysis. Am J Epidemiol 125:929–938.[Abstract/Free Full Text]

  4. Jarup L, Pershagen G, Wall S. (1989) Cumulative arsenic exposure and lung-cancer in smelter workers—a dose-response study. Am J Ind Med 15:31–41.[Web of Science][Medline]

  5. UNSCEAR. (2000) Sources and Effects of Ionising Radiation(United Nations, New York).

  6. Thun M, Schnorr T, Smith A, Halperin W, Lemen R. (1984) Cadmium and lung-cancer. Am J Epidemiol 120:482–483.

  7. Thun MJ, Schnorr TM, Smith AB, Halperin WE, Lemen RA. (1985) Mortality among a cohort of United States cadmium production workers—an update. J Natl Cancer Inst 74:325–333.[Web of Science][Medline]

  8. Lamm SH, Parkinson M, Anderson M, Taylor W. (1992) Determinants of lung cancer risk among cadmium-exposed workers. Ann Epidemiol 2:195–211.[Medline]

  9. Sorahan T and Lancashire RJ. (1997) Lung cancer mortality in a cohort of workers employed at a cadmium recovery plant in the United States: an analysis with detailed job histories. Occup Environ Med 54:194–201.[Abstract/Free Full Text]

  10. Sorahan T, Lister A, Gilthorpe MS, Harrington JM. (1995) Mortality of copper-cadmium alloy workers with special reference to lung-cancer and nonmalignant diseases of the respiratory system, 1946 –92. Occup Environ Med 52:804–812.[Abstract/Free Full Text]

  11. Jarup L, Bellander T, Hogstedt C, Spang G. (1998) Mortality and cancer incidence in Swedish battery workers exposed to cadmium and nickel. Occup Environ Med 55:755–759.[Abstract/Free Full Text]

  12. Sorahan T and Esmen NA. (2004) Lung cancer mortality in UK nickel-cadmium battery workers, 1947 –2000. Occup Environ Med 61:108–116.[Abstract/Free Full Text]

  13. Englyst V, Lundstrom NG, Gerhardsson L, Rylander L, Nordberg G. (2001) Lung cancer risks among lead smelter workers also exposed to arsenic. Sci Total Environ 273:77–82.[CrossRef][Medline]

  14. Jones RD. (1994) Survey of antimony workers: mortality 1961 –1992. Occup Environ Med 51:772–776.[Abstract/Free Full Text]

  15. Leonard A and Gerber GB. (1996) Mutagenicity, carcinogenicity and teratogenicity of antimony compounds. Mutat Res 366:1–8.[Web of Science][Medline]

  16. Schnorr TM, Steenland K, Thun MJ, Rinsky RA. (1995) Mortality in a cohort of antimony smelter workers. Am J Ind Med 27:759–770.[Web of Science][Medline]

  17. Holliday B. (1979) The management of plutonium workers. Radiol Prot Bull 27:22–26.

  18. Kravchick T, German U, Laichter Y, Weiser G. (1997) Estimate of Worker Intakes of Radioactive Materials Based on Air Sampling—A Review(Israel Atomic Energy Commission, Beersheba, Israel) Report No. N 97/606.

  19. Ogden TL, Bartlett IW, Purnell CJ, Wells CJ, Armitage F, Wolfson H. (1993) Dust from cotton manufacture: changing from static to personal sampling. Ann Occup Hyg 37:271–285.[Abstract/Free Full Text]

  20. Watt M, Godden D, Cherrie J, Seaton A. (1995) Individual exposure to particulate air pollution and its relevance to thresholds for health effects: a study of traffic wardens. Occup Environ Med 52:790–792.[Abstract/Free Full Text]

  21. Mulhausen JR and Daniano J. (1998) A Strategy for Assessing and Managing Occupational Exposures 2nd edn (American Industrial Hygiene Association Press, Fairfax, VA).

  22. Cohen AC. (1961) Tables for maximum likelihood estimates—singly truncated and singly censored samples. Technometrics 3:535–541.[CrossRef][Web of Science]

  23. Jones SR and Atkin P. (2007) Occupational Exposures at a Complex Non-Ferrous Metal Smelter, Humberside, UK(Westlakes Research Institute, Moor Row, Cumbria) Available from http://www.westlakes.org/live/html/pub/books.htm.

  24. Baxter MS, East BW, MacKenzie AB, Scott EM. (1990) A Review of Radioactivity in and Around the Capper Pass Smelter, Melton Works, North Humberside(Scottish Universities Research Reactor Centre, East Kilbride, UK).

  25. Richardson DB and Loomis D. (2004) The impact of exposure categorisation for grouped analyses of cohort data. Occup Environ Med 61:930–935.[Abstract/Free Full Text]

  26. ONS. (2003) Twentieth Century Mortality: 100 Years of Mortality Data in England and Wales by Age, Sex, Year and Underlying Cause(Mortality Statistics Unit, Office for National Statisitics, London, UK).

  27. Cytel. (1999) EGRET for Windows—Software for Analysis of Biomedical and Epidemiological Studies(Cytel Software Corporation, Cambridge, MA).

  28. Lubin JH, Blot WJ, Berrino F, et al. (1984) Modifying risk of developing lung cancer by changing habits of cigarette smoking. Br Med J (Clin Res Ed) 288:1953–1956.[Medline]

  29. Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. (2000) Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. Br Med J 321:323–329.[Abstract/Free Full Text]

  30. Wakai K, Seki N, Tamakoshi A, et al. (2001) Decrease in risk of lung cancer death in males after smoking cessation by age at quitting: findings from the JACC study. Jpn J Cancer Res 92:821–828.[CrossRef][Web of Science]

  31. Enterline PE and Marsh GM. (1980) Mortality studies of smelter workers. Am J Ind Med 1:251–259.[CrossRef][Medline]

  32. Moolgavkar SH and Luebeck G. (1990) Two-event model for carcinogenesis: biological, mathematical, and statistical considerations. Risk Anal 10:323–341.[CrossRef][Web of Science][Medline]

  33. Doll R. (1971) The age distribution of cancer: implications for models of carcinogenesis. J R Soc Med 134:Ser A133–166.

  34. Armitage P and Doll R. (2004) The age distribution of cancer and a multi-stage theory of carcinogenesis. Br J Cancer 91:1983–1989.[CrossRef][Web of Science][Medline]

  35. Moolgavkar SH and Knudson AG Jr. (1981) Mutation and cancer: a model for human carcinogenesis. J Natl Cancer Inst 66:1037–1052.[Web of Science][Medline]

  36. BEIR. (1999) Health Effects of Exposure to Radon (BEIR VI)(National Academy Press, Washington, DC).

  37. ICRP. (2001) Database of Dose Coefficients: Workers and Members of the Public (CD-ROM)(International Commission on Radiological Protection, Stockholm, Sweden).

  38. Lee-Feldstein A. (1986) Cumulative exposure to arsenic and its relationship to respiratory cancer among copper smelter employees. J Occup Med 28:296–302.[Web of Science][Medline]

  39. McCurdy SA, Sunyer J, Zock JP, Anto JM, Kogevinas M. (2003) Smoking and occupation from the European Community Respiratory Health Survey. Occup Environ Med 60:643–648.[Abstract/Free Full Text]

  40. Levin LI, Silverman DT, Hartge P, Fears TR, Hoover RN. (1990) Smoking patterns by occupation and duration of employment. Am J Ind Med 17:711–725.[Web of Science][Medline]

  41. Amandus H and Costello J. (1991) Silicosis and lung cancer in U.S. metal miners. Arch Environ Health 46:82–89.[Web of Science][Medline]

  42. Danielsen TE, Langard S, Andersen A, Knudsen O. (1993) Incidence of cancer among welders of mild-steel and other shipyard workers. Br J Ind Med 50:1097–1103.[Web of Science][Medline]

  43. Steenland K and Sanderson W. (2001) Lung cancer among industrial sand workers exposed to crystalline silica. Am J Epidemiol 153:695–703.[Abstract/Free Full Text]

  44. Steenland K. (2002) Ten-year update on mortality among mild-steel welders. Scand J Work Environ Health 28:163–167.[Web of Science][Medline]

  45. Doll R, Peto R, Boreham J, Sutherland I. (2005) Mortality from cancer in relation to smoking: 50 years observations on British doctors. Br J Cancer 92:426–429.[Web of Science][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
57/4/238    most recent
kql153v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Jones, S.
Right arrow Articles by Walker, P
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Jones, S.
Right arrow Articles by Walker, P
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?