IJE Advance Access originally published online on July 31, 2007
International Journal of Epidemiology 2007 36(5):1126-1135; doi:10.1093/ije/dym138
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mortality from diseases other than cancer following low doses of ionizing radiation: results from the 15-Country Study of nuclear industry workers
1International Agency for Research on Cancer, Lyon, France.
2Radiation Protection Bureau, Health Canada, Ottawa, Canada. Currently at McLaughlin Centre for Population Health Risk Assessment, University of Ottawa, Ottawa, Canada.
3University of Tampere, Tampere and STUK - Radiation and Nuclear Safety Authority, Helsinki, Finland.
4Department of Preventive Medicine, Cheju National University College of Medicine, Chejudo, Korea.
5The Nuclear Research Centre (SCK.CEN), Radiation Protection Division, Mol, Belgium.
6Radiation Epidemiology Branch, Division of Epidemiology and Genetics, National Cancer Institute, Bethesda, Maryland, USA.
7Department of Hygiene and Epidemiology, Faculty of Health Care and Social Work, Trnava University, Department of Hygiene and Epidemiology, Trnava, Slovak Republic.
8Faculty of Health Sciences, American University of Beirut, Lebanon.
9Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, USA.
10Lithuanian Cancer Registry, Vilnius University Oncology Institute, Vilnius, Lithuania.
11Midsweden Research and Development Center, Sundsvall Hospital, Sundsvall, Sweden.
12Radiation Protection Division, Health Protection Agency, Chilton, Didcot, UK.
13Department of Epidemiology, School of Public Health, Chapel Hill, NC, USA.
14Department of Preventive Medicine and Public Health, School of Medicine, Universidad Autonoma de Madrid, Spain.
15Institut de Radioprotection et de Sûreté Nucléaire, Fontenay-aux-Roses, France.
16National Institute for Occupational Safety and Health, Cincinnati, USA.
17Office Cantonal de l'Inspection et des Relations du Travail, Genève, Suisse.
18Department of Dermatology, Venereology and Dermatooncology, Semmelweis University, Budapest, Hungary.
Deceased.
*Corresponding author. International Agency for Research on Cancer, 150, Cours Albert Thomas, 69372 Lyon Cedex 08, France. E-mail: vrijheid{at}iarc.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
Background Ionizing radiation at very high (radio-therapeutic) dose levels can cause diseases other than cancer, particularly heart diseases. There is increasing evidence that doses of the order of a few sievert (Sv) may also increase the risk of non-cancer diseases. It is not known, however, whether such effects also occur following the lower doses and dose rates of public health concern.
Methods We used data from an international (15-country) nuclear workers cohort study to evaluate whether mortality from diseases other than cancer is related to low doses of external ionizing radiation. Analyses included 275 312 workers with adequate information on socioeconomic status, over 4 million person-years of follow-up and an average cumulative radiation dose of 20.7 mSv; 11 255 workers had died of non-cancer diseases.
Results The excess relative risk (ERR) per Sv was 0.24 [95% CI (confidence intervals) –0.23, 0.78] for mortality from all non-cancer diseases and 0.09 (95% CI –0.43, 0.70) for circulatory diseases. Higher risk estimates were observed for mortality from respiratory and digestive diseases, but confidence intervals included zero. Increased risks were observed among the younger workers (attained age <50 years, identified post hoc) for all groupings of non-cancer causes of death, including external causes. It is unclear therefore whether these findings reflect real effects of radiation, random variation or residual confounding.
Conclusions The most informative low-dose radiation study to date provides little evidence for a relationship between mortality from non-malignant diseases and radiation dose. However, we cannot rule out risks per unit dose of the same order of magnitude as found in studies at higher doses.
Keywords Radiation, ionizing, radiation effects, cohort studies, occupational diseases, cardiovascular diseases, digestive system diseases, respiratory tract diseases
Accepted 6 June 2007
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
Ionizing radiation is a well-established human carcinogen. At high doses, ionizing radiation exposure is also known to cause non-cancer effects. Excess cardiovascular mortality has been observed in studies of patients treated with radiotherapy for breast cancer1–4 and Hodgkin's disease5 who received doses of 10 sievert (Sv) or more to the heart at high dose-rates (for simplicity, all doses in this article are expressed in terms of equivalent dose in Sv to a particular tissue (colon, lung). As this article is mainly concerned with photon radiation with a radiation-weighting factor of 1, the results could equally well be expressed in terms of absorbed dose to organs in grays with the same numerical values6). Very high doses of radiation to the heart can result in fibrotic and small vessel damage in the pericardium, myocardium and coronary vasculature, with subsequent myocardial infarction being the most common fatal complication.7
Non-cancer late effects have been documented at lower doses among the atomic bomb survivors.8–12 Radiation-related excess mortality and morbidity due to circulatory, digestive and respiratory diseases have been observed in this cohort at doses below 4 Sv; however, a threshold as high as 500 mSv could not be ruled out.11 Other studies of populations with doses below 5 Sv have, to date, not shown a consistent pattern of increases in mortality from circulatory disease (reviewed by McGale and Darby),13 although a dose-related increase in coronary heart disease was observed recently in patients receiving radiotherapy for peptic ulcer with cardiac doses ranging from 1.6 to 3.9 Sv.14 Further, a recent study among Chernobyl clean-up workers, with an average whole-body dose of 100 mSv, reported an increased risk of ischaemic heart disease, essential hypertension and cerebrovascular disease; analyses were, however, not adjusted for smoking or alcohol consumption.15
Given the nature of the radiation damage to the heart and vessels observed following radiotherapy,7 it is not known whether any increased risk is in fact expected following the low doses (of the order of a few tens of mSv) received in a protracted or fractionated fashion that are of public health concern. If such effects were confirmed, this would have important implications for radiation protection of medically, occupationally and environmentally exposed-populations, as well as for the understanding of the aetiology of circulatory and other non-cancer diseases.
Workers in the nuclear industry are a suitable population for the direct estimation of radiation effects at low doses: they form large, relatively stable populations with relatively well-measured and well-recorded external radiation doses. Large studies of combined cohorts of nuclear workers have previously reported no radiation-related excess in mortality from non-cancer diseases as a group.16,17 A dose-related increased mortality from circulatory diseases has been observed in some studies of nuclear industry workers,17,18 but it is unclear whether this reflects a real effect of radiation exposure or residual confounding.
This article evaluates whether mortality from diseases other than cancer is related to radiation dose in the 15-Country Study of nuclear industry workers.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
The 15-Country Study is a multi-national retrospective cohort study, in which the mortality of nuclear industry workers from 15 countries was studied. In this article, the term nuclear industry is used to refer to facilities engaged in the production of nuclear power, the manufacture of nuclear weapons, the enrichment and processing of nuclear fuel, the production of radioisotopes or reactor or weapons research; uranium mining is not included. Radiation-related cancer risk estimates from the 15-Country Study have been published elsewhere,19,20 as have the detailed methods of the follow-up20 and dosimetry.22 The main study population was defined21 as workers who had been employed in at least one of the study facilities for at least 1 year, who had been monitored for external radiation exposure and whose doses resulted predominantly from exposure to higher energy photon radiation (X and
-rays in the range 100–3000 keV). For the non-cancer analyses presented here, workers also had to have been followed for non-cancer mortality (this excludes the Japanese cohort), and have adequate socioeconomic status (SES) information (this excludes the Japanese, US-Idaho National Laboratory and Canada-Ontario Hydro cohorts).21 For each worker monitored for external radiation exposure, individual annual recorded radiation doses were obtained from facility records and/or national dose registries. Doses to specific organs were derived by dividing the recorded doses by organ dose bias factors developed in a study of errors in doses.22 In this article, lung doses are used for respiratory diseases and colon doses for all other diseases, as was done in the study of atomic bomb survivors.11
Statistical analysis
Consistent with previous analyses of this cohort,20 doses and person-years at risk for each worker were accumulated over time from date of entry in the study to date of exit. To allow for a possible latent period between an exposure and its effect, cumulative doses were lagged by 10 years, for consistency with radiotherapy studies.4,7 Sensitivity analyses used a range of different lag times.
Observed (O) and expected (E) numbers of deaths and person-years of follow-up were calculated by eleven dose categories (0–, 5–, 10–, 20–, 50–, 100–, 150–, 200–, 300–, 400– and 500+ mSv) chosen a priori. The expected numbers of deaths were calculated assuming that, within a stratum defined by levels of the confounding variables, the mortality rate in each dose category was the same as that of the entire stratum, i.e. that the cause of death under study was not associated with exposure. The score test statistic based on the linear relative risk model was used to test the statistical significance of the trend in O/E ratios.23
Analyses were based on a linear relative risk Poisson model, in which the relative risk is of the form 1 + ßZ, where Z is the lagged cumulative dose in Sv, and ß is the excess relative risk (ERR) per sievert. This model has been used commonly in analyses of nuclear workers studies16,17 and radiation risk estimation.6,24 The linear RR model has computational restrictions, however, as the relative risk cannot be negative. Log-linear models, in which the relative risk (RR) is assumed to be of the form exp(ßZ), were therefore also fitted to the data. Resulting estimates of the RR at a dose of 100 mSv compared with 0 mSv are presented where ß could not be estimated under the linear model. Linear RR and log-linear models give essentially the same results at low doses and low risks. Confidence intervals are 95% likelihood-based intervals.
All analyses (trend tests and regression models) were stratified on sex, age and calendar period (both in 5-year categories), facility, duration of employment (<10 years, 
10 years) and SES.
The risk of radiation-induced disease may vary among subgroups of the study population (gender, study-cohort, facility type) and by variables such as attained age, age at exposure and time since exposures.11 Heterogeneity of risk estimates was tested using likelihood ratio tests for the interaction of gender, cohort, facility type and age with dose. A priori defined categorical variables were used to evaluate variation in risk by attained age (<60, 60–70, 70+ years), age-at-exposure (<35, 35–50 and 50+ years) and time-since-exposure (<10, 10–20 and 20+ years). For age-at-exposure and time-since-exposure, the cumulative doses received in each of the three categories were modelled jointly, and the likelihood of this model was tested against that of the standard model with one total cumulative dose.
Outcomes studied
Results are presented for mortality from all non-malignant (non-cancer) diseases combined and for the main groupings of specific non-cancer diseases: circulatory diseases, respiratory diseases excluding pneumonia and digestive diseases. Particular attention is paid to circulatory diseases, because of the evidence from high-dose studies discussed above. Mortality from chronic obstructive pulmonary disease, liver cirrhosis and external causes were also studied as indirect indicators of potential for confounding by, respectively, smoking, alcohol consumption and more general life-style related factors. The ICD codes defining these causes of death have been published elsewhere.21 Only underlying causes of death were analysed. Because of limited statistical power, analyses of age-at-exposure and time-since-exposure were carried out only for all non-cancer diseases combined and for circulatory diseases.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
The study population for the non-cancer analyses included 275 312 workers with a total duration of follow-up of 4 067 861 person-years and a total collective recorded dose of 5690 Sv (Table 1). The average cumulative recorded dose was 20.7 mSv; <0.1% of workers had received cumulative doses >500 mSv over their entire career in the nuclear industry.19 The average individual colon dose and lung dose were 16.9 and 17.9 mSv, respectively.
|
A total of 13 315 workers died from causes other than cancer, 2060 from external causes of death and 11 255 from non-cancer diseases; the majority of these latter deaths were from circulatory diseases (n = 8412) (Table 2). The ERR per sievert for mortality from all non-cancer diseases was 0.24 (95% CI –0.23, 0.78) (Table 2). The corresponding trend test gave little evidence for an increase in risk with dose (P = 0.16). There was no evidence of a dose–response for circulatory disease mortality overall (ERR/Sv = 0.09; 95% CI –0.43, 0.70; P for trend = 0.37) or for mortality from specific circulatory diseases. Elevated ERRs/Sv were observed for cerebrovascular diseases (ERR/Sv = 0.88; 95% CI –0.67, 3.16) and for the category ‘other circulatory diseases’ (ERR/Sv = 0.29; 95% CI < 0, 2.40), but confidence intervals included zero.
|
Risk estimates for mortality from all non-malignant respiratory diseases (ERR/Sv = 1.16; 95% CI –0.53, 3.84), chronic obstructive pulmonary disease (ERR/Sv = 0.56; 95% CI < 0, 3.77), all digestive diseases (ERR/Sv = 0.96; 95% CI < 0, 4.52) and liver cirrhosis (ERR/Sv = 1.54; 95% CI < 0, 9.67) were elevated but confidence intervals included zero. External causes of death showed a negative ERR/Sv (–0.50, 95% CI < 0, 1.08) (Table 2).
There was little evidence for radiation-related increases in mortality from non-cancer diseases overall or from specific disease groups in individual countries, or for heterogeneity of risks between countries, cohorts or facility types (Supplementary Table A). However, numbers of deaths were small in many countries and the confidence intervals were wide. Elevated risks were, nevertheless, seen in the US-NPP cohort, for all non-cancer diseases (ERR/Sv = 5.99; 95% CI 1.75, 12.4) and for circulatory diseases (ERR/Sv = 6.04; 95% CI 1.36, 13.5).
The ERRs/Sv in men were very similar to those observed in the entire population for mortality from all non-cancer diseases as well as from circulatory, respiratory and digestive diseases (Table 3; Supplementary Table B), reflecting the fact that the study population consisted mainly of men (90%) who received 98% of the collective dose. There was no evidence for a difference in risk between men and women.
|
There was some evidence for heterogeneity of risk by attained age category for non-cancer diseases as a group although confidence intervals for the different age groups overlapped (heterogeneity P = 0.10); the ERR/Sv was highest for attained ages below 60 years (ERR/Sv = 1.25; 95% CI –0.03, 2.83) (Table 3). More detailed analyses conducted post hoc of attained age in four categories (Table 4) showed that the elevated risks were found mainly in the age group below 50 years, and that the increased risk in this age category was seen for each grouping of non-cancer causes of death, including external causes.
|
There was no evidence for a difference in radiation-related risk by age at exposure for all non-cancer diseases or for circulatory diseases (Table 3). Differences in ERR/Sv estimates were found, however, between time-since-exposure categories for both of these outcomes; this was due to negative ERRs/Sv for doses received <10 years previously (Table 3). Risk estimates tended to increase with increasing lag time for all non-cancer diseases overall, as well as for circulatory, respiratory and digestive diseases (Table 3; Supplementary Table B).
ERRs/Sv increased from 0.24 to 0.48 for all non-cancer diseases and from 0.09 to 0.39 for circulatory disease when analyses were not stratified for SES (Table 3). Removing stratification for duration of employment resulted in a reduction in estimated ERRs/Sv from 0.24 to 0.04 for all non-cancer diseases, and from 0.09 to –0.04 for circulatory diseases. Similar results were observed for respiratory and digestive diseases (Supplementary Table B).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
The 15-Country Study gives little evidence for a relation between mortality from non-malignant diseases and external radiation dose at the low doses received by nuclear industry workers. However, radiation-related non-cancer effects are expected to be small (if they exist at all) at the low doses received by most nuclear workers. Therefore, despite the large size of this study, the power to detect a small increase in risk is limited. Indeed, the central risk estimates observed in this study for mortality from all non-cancer diseases and from circulatory diseases are similar to comparable estimates derived from the A-bomb survivors’ data (Table 5), and thus statistically compatible with a linear extrapolation from these data. Hence, we can exclude neither the existence of a threshold for the induction of non-cancer effects, nor an effect of the same order of magnitude as found in the A-bomb survivors. These findings are consistent with most other low-dose studies, in which radiation-related excess in risk of circulatory disease have generally not been observed (reviewed by McGale and Darby).13
|
Larger increases in risk were observed for mortality from respiratory diseases and digestive diseases in this study. The A-bomb survivors’ study has reported radiation-related increases in risk for these disease outcomes,11 but, to our knowledge, no other large studies have reported such associations. The explanation for these results is unclear and may include chance (given the wide confidence intervals), as well as confounding by factors such as smoking and alcohol consumption.
Analyses in this study are not independent from those of the previous 3-Country Study which combined cohorts from Canada, the UK and the US.17 The weakly positive risk estimates in this study were driven mainly by the cohorts and follow-up periods also included in the 3-Country Study (3-country cohorts: ERR/Sv for non-cancer diseases = 0.40; 95% CI –0.18, 1.08. Other cohorts: ERR/Sv = –0.11; 95% CI < 0, 0.85). These are also among the cohorts with the longest follow-up periods and the largest numbers of workers who accumulated higher doses.
Time and age-related effects
Of interest is the dose–response observed for all groupings of non-cancer diseases, including external causes, among workers with younger attained ages (below 50 years of age). This result is consistent with the A-bomb survivors’ data where a decrease in mortality from all non-cancers with increasing attained age is observed.11 It is unclear, however, whether our finding reflects a real effect of radiation, random variation or residual confounding, particularly since it is not related to any particular cause of death and since this age group was identified post hoc. Given the elevated risk for external causes in the younger age group, a confounding effect of life-style related factors acting particularly at younger ages appears possible.
The current study finds no evidence for differences in risk for radiation doses received at different ages. The A-bomb survivors’ study reports a trend of decreasing ERR with increasing age at exposure for all non-cancer diseases combined,11 whereas a more recent re-analysis of these same data has indicated that the highest risk arose in the medium (30–49 years) age group.25
Analyses by time-since-exposure and different lag periods showed higher risk estimates for exposures 10 years or more in the past. These findings are consistent with higher-dose studies in which radiation-related effects were generally not seen until at least 10 years after exposure.4,7 A-bomb survivors analyses have now documented non-cancer effects in the follow-up periods covering 
20–50 years after exposure.11,25
Confounding
A limitation of this study is that, as in most occupational cohort studies, information on factors such as smoking habits, alcohol intake and diet could not be obtained. Non-cancer disease risks are known to be strongly related to such risk factors.26 Smoking surveys in seven of the study cohorts have shown no consistent association between radiation dose and smoking with three showing a positive association and four showing no association.21 Nevertheless, lung cancer risks were particularly raised in the 15-Country Study19,20 suggesting that confounding by smoking may have played a role; a similar confounding effect in the analyses of non-cancer disease cannot be ruled out. The increased risk estimate for chronic obstructive pulmonary diseases, although its confidence interval includes zero, is compatible with this. Further, it is notable that risk estimates for respiratory diseases and liver cirrhosis in the 15-Country Study are much higher than comparable estimates from the A-bomb survivors (Table 5), possibly demonstrating confounding effects by, respectively, smoking and alcohol consumption. Increases observed in the younger attained age categories for all main causes of non-cancer disease as well as external mortality causes may also indicate a specific confounding effect of lifestyle-related risk factors at younger ages.
This study partially controlled for these effects through adjustment for SES. SES was found to be an important confounder in this study, as it was associated both with all cause mortality and exposure in most countries,21 and removal of SES stratification in our analyses had a relatively large effect of increasing risk estimates. Nevertheless, a surrogate measure can never entirely compensate for lack of direct data, and confounding may explain some of the small increases in radiation-related risk reported for non-cancer diseases.
Duration of employment was considered a priori to be a possible confounder and was therefore included in all main analyses to control for the so-called ‘healthy worker survivor effect’.20 This negative confounding effect occurs when workers who are healthier and therefore have lower mortality rates, stay in employment longer and may accumulate higher doses.27 When we removed adjustment for duration of employment in the analyses, ERR estimates reduced to below 0, indicating that a healthy worker survivor effect may be present in this cohort.
Outcome misclassification
Misclassification of vital status or of underlying cause of death can occur in mortality studies. Coverage of ascertainment of both vital status and cause of death were near complete in the study cohorts, limiting the potential of bias.21 Misclassification of the cause of death recorded on death certificates is likely to have occurred, especially for specific causes of death,28–31 and may have differed over time and between countries. Nevertheless, risk estimates for the group of all non-cancer diseases combined would remain relatively unaffected.
Errors in dosimetry
Nuclear workers studies benefit from relatively well-measured and well-recorded radiation doses. However, even these doses can be subject to errors. An extensive study of errors in dosimetry formed part of the 15-Country Study.22 This study identified and quantified the main errors in recorded doses, and dose estimates in the main analyses have been adjusted to take these errors into account. Further, analyses were restricted to workers whose radiation dose was predominantly from relatively well-measured radiation types (external radiation from higher energy photons), and workers with substantial doses from other radiation types (neutrons, internal contamination) were excluded.22
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
In conclusion, results from the largest nuclear workers cohort study to date give little evidence for a relation between mortality from non-malignant diseases and external radiation dose at the low doses received by nuclear industry workers. However, we cannot rule out risks of the same order of magnitude as found in the A-bomb survivors’ study. Dose–response relationships found in younger workers merit further exploration. The follow-up periods in most countries were relatively short (average 13 years) and the majority of workers were comparatively young at the end of the follow-up (average 46 years). Further follow-up of these populations will be useful to improve the precision of risk estimates.
|
Supplementary material |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
Supplementary data are available at IJE Online.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
The authors are grateful to the late Sir Richard Doll and to Prof. Jacques Estève and Prof. Bruce Armstrong who helped to start this study; to all persons in the participating countries who worked on the data collection and validation; and to the representatives and staff of the nuclear facilities included in the study for their open collaboration.
Financial support for coordination of the International Study was provided by the European Union (contracts F13P-CT930066, F14P-CT96-0062, FIGH-CT1999-20001), the US Centers for Disease Control and Prevention (Co-operative agreement U50/CCU011778) and the Canadian Nuclear Safety Commission. These sponsors had no role in study design, data collection, analysis or interpretation. Funding was also obtained from one of the collaborating institutes: the Japanese Institute for Radiation Epidemiology.
Funding sources for the national studies included: Australian Nuclear Science and Technology Organisation; Nuclear Research Center (SCK.CEN), Belgium; Health Canada and Statistics Canada; La Ligue Nationale contre le Cancer, France; La Compagnie Générale des Matière Nucléaire, France; Electricité de France; Ministry of Education, Culture, Sports, Science and Technology in Japan; Ministry of Science & Technology (MOST) in Korea; Spanish Nuclear Safety Council; the UK Health and Safety Executive; the US Department of Energy.
The study was approved by the IARC Ethical Review Committee and by the relevant ethics committees of the participating countries. The procedures followed were in accordance with the ethical standards of the responsible committees on human experimentation (institutional or regional) and with the Helsinki Declaration32 (revision depending on the country). The study did not involve contact with study subjects.
Conflict of interest: None declared.
KEY MESSAGES
|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsSupplementary materialAcknowledgementsReferences |
|---|
1 Darby S, McGale P, Peto R, Granath F, Hall P, Ekbom A. Mortality from cardiovascular disease more than 10 years after radiotherapy for breast cancer: nationwide cohort study of 90 000 Swedish women. Br Med J (2003) 326:256–57.
2 Early Breast Cancer Trialists’ Collaborative Group. Favourable and unfavourable effects on long-term survival of radiotherapy for early breast cancer: an overview of the randomised trials. Lancet (2000) 355:1757–70.[CrossRef][Web of Science][Medline]
3 Paszat LF, Mackillop WJ, Groome PA, Boyd C, Schulze K, Holowaty E. Mortality from myocardial infarction after adjuvant radiotherapy for breast cancer in the surveillance, epidemiology, and end-results cancer registries. J Clin Oncol (1998) 16:2625–31.[Abstract]
4 Darby SC, McGale P, Taylor CW, Peto R. Long-term mortality from heart disease and lung cancer after radiotherapy for early breast cancer: prospective cohort study of about 300,000 women in US SEER cancer registries. Lancet Oncol (2005) 6:557–65.[CrossRef][Web of Science][Medline]
5 Hancock SL, Tucker MA, Hoppe RT. Factors affecting late mortality from heart disease after treatment of Hodgkin. JAMA (1993) 270:1949–55.
6 ICRP (International Commission on Radiological Protection). Recommendations of the International Commission on Radiological Protection; ICRP Report 60: Publication 60. Ann ICRP (1991) 49:1–3.
7 Gaya AM, Ashford RF. Cardiac complications of radiation therapy. Clin Oncol (R Coll Radiol) (2005) 17:153–59.[Medline]
8 Wong FL, Yamada M, Sasaki H, et al. Noncancer disease incidence in the atomic bomb survivors: 1958-1986. Radiat Res (1993) 135:418–30.[Web of Science][Medline]
9 Kodama K, Fujiwara S, Yamada M, Kasag F, Shimizu Y, Shigematsu I. Profiles of non-cancer diseases in atomic bomb survivors. World Health Stat Q (1996) 49:7–16.[Medline]
10 Shimizu Y, Pierce D, Preston DL, Mabuchi K. Studies of the mortality of atomic bomb survivors. Report 12, part II. Noncancer mortality: 1950-1990. Radiat Res (1999) 152:374–89.[Web of Science][Medline]
11 Preston DL, Shimizu Y, Pierce DA, Suyama A, Mabuchi K. Studies of mortality of atomic bomb survivors. Report 13: solid cancer and noncancer disease mortality: 1950–1997. Radiat Res (2003) 160:381–407.[CrossRef][Web of Science][Medline]
12 Yamada M, Wong F, Fujiwara S, Akahoshi M, Suzuk G. Noncancer disease incidence in atomic bomb survivors, 1958–1998. Radiat Res (2004) 161:622–32.[CrossRef][Web of Science][Medline]
13 McGale P, Darby SC. Low doses of ionizing radiation and circulatory diseases: a systematic review of the published epidemiological evidence. Radiat Res (2005) 163:247–57.[CrossRef][Web of Science][Medline]
14 Carr ZA, Land CE, Kleinerman RA, et al. Coronary heart disease after radiotherapy for peptic ulcer disease. Int J Radiat Oncol Biol Phys (2005) 61:842–50.[CrossRef][Web of Science][Medline]
15 Ivanov VK, Maksioutov MA, Chekin SY, et al. The risk of radiation-induced cerebrovascular disease in Chernobyl emergency workers. Health Phys (2006) 90:199–207.[CrossRef][Web of Science][Medline]
16 Muirhead CR, Goodill AA, Haylock RGE, et al. Occupational radiation exposure and mortality: second analysis of the National Registry for Radiation Workers. J Radiol Prot (1999) 19:3–26.[CrossRef][Medline]
17 Cardis E, Gilbert ES, Carpenter L, et al. Effects of low doses and low dose rates of external ionizing radiation: cancer mortality among nuclear industry workers in three countries. Radiat Res (1995) 142:117–32.[Web of Science][Medline]
18 Ashmore JP, Krewski D, Zielinski JM, Jiang H, Semenciw R, Band PR. First analysis of mortality and occupational radiation exposure based on the National Dose Registry of Canada. Am J Epidemiol (1998) 148:564–74.
19 Cardis E, Vrijheid M, Blettner M, et al. Risk of cancer after low doses of ionising radiation: retrospective cohort study in 15 countries. Br Med J (2005) 331:77–80.
20 Cardis E, Vrijheid M, Blettner M, et al. The 15-Country Collaborative Study of cancer risk among radiation workers in the nuclear industry: estimates of radiation related cancer risks. Radiat Res (2007) 167:396–416.[CrossRef][Web of Science][Medline]
21 Vrijheid M, Cardis E, Blettner M, et al. The 15-Country Collaborative Study of cancer risk among radiation workers in the nuclear industry: design, epidemiological methods and descriptive results. Radiat Res (2007) 167:361–79.[CrossRef][Web of Science][Medline]
22 Thierry-Chef I, Marshall M, Fix JJ, et al. The 15-Country Collaborative Study of cancer risk among radiation workers in the nuclear industry: study of errors in dosimetry. Radiat Res (2007) 167:380–95.[CrossRef][Web of Science][Medline]
23 Hakulinen T. A Mantel-Haenszel statistic for testing the association between a polychotomous exposure and a rare outcome. Am J Epidemiol (1981) 113:192–97.
24 US NRC. Committee on the Biological Effects of Ionizing Radiation. (2006) Health risks from exposures to low levels of ionizing radiation, BEIR VII.
25 Zhang W, Muirhead CR, Hunter NJ. Age-at-exposure effects on risk estimates for non-cancer mortality in the Japanese atomic bomb survivors. Radiol Prot (2005) 25:393–404.[CrossRef]
26 Evans A, Tolonen H, Hense H-W, Ferrario M, Sans S, Kuulasmaa K, for the WHO MONICA Project. Trends in coronary risk factors in the WHO MONICA Project. Int J Epidemiol (2001) 30:S40.
27 McNamee R. Confounding and confounders. Occup Environ Med (2003) 60:227–34.
28 Chen L, Walker S, Tong S. The impact of the variation in death certification and coding practices on trends in mortality from ischaemic heart disease. Austr Health Rev (2002) 25:189–97.
29 Johansson LA, Westerling R. Comparing Swedish hospital discharge records with death certificates: implications for mortality statistics. Int J Epidemiol (2000) 29:495–502.
30 Lloyd-Jones DM, Martin DO, Larson MG, Levy D. Accuracy of death certificates for coding coronary heart disease as the cause of death. Ann Intern Med (1998) 129:1020–6.
31 Benavides FG, Bolumar F, Peris R. Quality of death certificates in Valencia, Spain. Am J Public Health (1989) 79:1352–4.
32 World Medical Association Declaration of Helsinki. Recommendations Guiding Medical Doctors in Biomedical Research Involving Human Subjects, 1983. Amended by the 35th World Medical Assembly.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M P Little, E J Tawn, I Tzoulaki, R Wakeford, G Hildebrandt, S Tapio, and P Elliott Comments: The non-cancer mortality experience of male workers at British Nuclear Fuels plc, 1946-2005 Int. J. Epidemiol., August 1, 2009; 38(4): 1159 - 1164. [Full Text] [PDF]
|
|
|
|
|
|
D. McGeoghegan, K. Binks, M. Gillies, S. Jones, and S. Whaley The non-cancer mortality experience of male workers at British Nuclear Fuels plc, 1946-2005 Int. J. Epidemiol., June 1, 2008; 37(3): 506 - 518. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||