Occupational Medicine Advance Access originally published online on August 29, 2007
Occupational Medicine 2007 57(7):518-524; doi:10.1093/occmed/kqm078
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Occupational exposure to ionizing and non-ionizing radiation and risk of glioma
1 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Victoria 3800, Australia
2 Non-Ionising Radiation Branch, Australian Radiation Protection and Nuclear Safety Agency, Yallambie, Victoria 3085, Australia
3 Centre of Expertise for Good Practices and Competence, Finnish Institute of Occupational Health, Helsinki 00250, Finland
4 Cancer Epidemiology Centre, The Cancer Council of Victoria, Carlton, Victoria 3053, Australia
Correspondence to: Ken K. Karipidis, 619 Lower Plenty Road, Yallambie, Victoria 3085, Australia. Tel: +61 3 9433 2282; fax: +61 3 9432 1835; e-mail: ken.karipidis{at}arpansa.gov.au
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
Background Although the aetiology of glioma is poorly understood, the higher incidence in males has long suggested an occupational cause.
Aim To investigate possible associations between occupational exposure to ionizing, ultraviolet (UV), radiofrequency (RF) and extremely low frequency (ELF) radiation and adult glioma risk.
Methods Case–control study using histologically confirmed cases of glioma first diagnosed between 1987 and 1991 in Melbourne, Australia, matched by age, sex and postcode of residence. A detailed occupational history was obtained for each subject. Exposure to radiation was assessed using a Finnish job exposure matrix (FINJEM) for all the radiation types as well as self-reports and expert hygienist review for RF and ionizing radiation. For ELF and UV, gender-specific FINJEM analysis was performed.
Results The study population consisted of 416 cases of glioma and 422 controls. The risk estimates given by FINJEM for ELF, RF and ionizing radiation were close to or below unity. Gender-specific analysis for UV showed odds ratios of 1.60 [95% confidence interval (CI) 0.95–2.69] and 0.54 (95% CI 0.27–1.07) for the highest exposed group of men and women, respectively (corresponding P value for trend was 0.03 and 0.04).
Conclusions We did not find evidence of an association between glioma and occupational exposure to ELF, RF and ionizing radiation. UV radiation was associated with increased glioma risk for men but this result could have been confounded by other predominantly male occupational and lifestyle exposures associated with high UV. Further investigation of UV radiation and glioma risk is suggested.
Keywords Cancer; case control; occupational exposure; radiation
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
Glioma is the most common primary malignant brain tumour in adults and has a poor prognosis. The age-specific incidence of this tumour is increasing in many developed countries, and this increase appears to be only partly explained by improved diagnostic techniques [1]. Although, several epidemiological studies have investigated the role of potential risk factors for glioma, its aetiology remains poorly understood, and there appears to be no association between glioma and the various dietary and lifestyle factors commonly associated with cancer at other body sites [2]. Previous analyses of dietary factors, smoking and alcohol consumption and medical, family and reproductive histories in our study generally did not find any associations with glioma risk [3–5].
The higher incidence of glioma in males has long suggested an occupational cause and possible risk factors include ionizing and non-ionizing radiation; the latter comprising of ultraviolet (UV), radiofrequency (RF) and extremely low frequency (ELF) radiation. Several previous studies have examined the possibility that occupational exposure to radiation increases the risk of brain tumour but their results have been largely inconsistent [6–8]. The majority of previous studies, however, have been limited by low statistical power, diagnostic non-specificity and poor assessment of exposure.
Our study is one of the largest of a series of 10 case–control studies within the Surveillance of Environmental Aspects Related to Cancer in Humans (SEARCH) program, which was co-ordinated by the International Agency for Research on Cancer in Lyon, France. We used a Finnish job exposure matrix (FINJEM) to investigate the relationship between occupational ionizing, UV, RF and ELF radiation exposure and adult glioma in an Australian community-based case–control study. Our aim was to determine whether occupational exposure to ionizing and non-ionizing radiation is associated with an increased risk of developing glioma.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
Information on the case and control ascertainment has been given in detail previously [3]. Cases were defined as people with histologically confirmed primary glioma (International Classification of Disease for Oncology codes 938–946) diagnosed between July 1987 and December 1991 [9], who were aged between 20 and 70 years at diagnosis and lived in Melbourne or four other major population centres in the state of Victoria, Australia. Cases were identified by screening the medical records of 14 Melbourne hospitals, which together provide most of the neurosurgical services within Victoria. Completeness of ascertainment was checked by reference to the Victorian Cancer Registry, to which notification of the incident cases of cancer is required by law.
Controls were frequency matched to cases by 5-year age groups and gender from a random sample of people listed on the Victorian electoral roll who lived in the areas from which the cases were recruited. Registration to vote is compulsory in the State of Victoria and the electoral roll contained 85% of all people who were eligible to vote in 1989 [10]. Exclusion criteria for potential controls included a history of stroke, epilepsy or brain tumour.
The research protocol was approved by the Standing Committee on Ethics in Human Research at Monash University, and informed consent was obtained from all subjects.
Initially, a self-administered questionnaire which included a work history booklet was mailed to the subjects to ascertain their occupational history. The booklet requested information about occupation, employer, industry, main tasks and duties, equipment used, start and finish dates, number of hours worked per day and number of days worked per week. Participants were requested to include all jobs they had held for at least 3 months since the age of 12 years. Two weeks after the mailed questionnaire, subjects were interviewed face-to-face by a research nurse. At the time of the interview, demographic details including age, sex, schooling and marital status were recorded. The information in the work history booklet was checked for completeness with the subject, and any gaps were filled. For 182 cases (44%), where the subject was unable to complete the questionnaires due to death or disability, this was completed by the subject's next of kin (proxy interview). There were only 2% of proxy interviews for the controls.
Exposure to ELF, RF, UV and ionizing radiation was estimated using the community-based job exposure matrix called FINJEM, developed by the Finnish Institute of Occupational Health [11]. FINJEM covers a wide range of occupational exposures and is the only job exposure matrix that covers all the different radiation types. Previous experience with the application of FINJEM has found it to be useful for some chemical exposures in the Australian workplace [12]. Exposure estimates in FINJEM are provided for five time periods between 1960 and 2003 (1960–84, 1985–94, 1995–97, 1998–2000 and 2001–03) and are based on industrial hygiene measurements, sample surveys and expert assessment.
Exposure in FINJEM is described by the proportion of exposed workers (P) and the annual mean level of exposure among the exposed (L); given as P x L for each occupation. The level of exposure is expressed in agent-specific units: µT for ELF, W/m2 for RF, J/m2 for UV and mSv for ionizing radiation. UV radiation exposure in FINJEM includes both natural (sunlight, regular outdoor work) and artificial (e.g. from welding). Occupational exposures to levels that are deemed not to exceed the non-occupational background are omitted. FINJEM classifies 74 occupations for ELF, 4 for RF, 102 for UV and 19 for ionizing radiation.
In order to assign the relevant radiation exposure according to FINJEM, all jobs held by each subject were first allocated to the relevant Finnish occupational codes. The exposure for each subject was then calculated by multiplying the duration of employment in each work history record by the corresponding entry of FINJEM (P x L), according to the time period of the exposure. If the exposure took place before 1960, FINJEM estimates for the period 1960–84 were used. Cumulative exposure estimates were then calculated by aggregating radiation exposure across the total work history in units of ‘radiation specific unit’ years (e.g. µT years for ELF).
For RF and ionizing radiation, exposure assessment included self-reports and expert hygienist review. During the interview, subjects were asked whether they had been exposed or not to RF and/or ionizing radiation for each job that they had and this constituted the ‘self-reported’ exposures. The expert review was conducted by an industrial hygienist with 15 years experience in assessing radiation exposures in workplaces in Victoria. The hygienist was blind to the case or control status of the subjects and to the self-reported exposures. The hygienist applied a dichotomous exposure classification based on probable high exposure to RF and/or ionizing radiation for each job. For both the self-reported and industrial hygienist exposure assessment methods, the duration of exposure in years for the jobs listed as being exposed was aggregated across the total work history.
All analyses were performed using SPSS software (SPSS Inc., Chicago, IL, USA). The analysis using FINJEM was performed for all the radiation types. Because of the higher incidence of glioma in males, separate FINJEM analysis for males and females was performed for ELF and UV (gender-specific analysis was not performed for RF and ionizing radiation because the numbers of exposed subjects were too low). Analyses using self-reports and expert assessment were performed only for RF and ionizing radiation due to lack of data for UV; results for ELF using self-reports and expert assessment are available elsewhere [13]. Proxy interviews were included in all analyses because separate analyses including only index subjects (i.e. excluding proxies) showed no major differences in the results (data not shown).
For all the exposure assessment methods, we compared four levels of exposure by using the unexposed subjects as the reference group and dividing the cumulative exposure (FINJEM) or the duration of exposure (self-reports/expert assessment) distributions of the exposed subjects into tertiles. Unconditional logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (95% CIs) adjusted for the matching variables: age (as a continuous variable) and sex. The model also included years of schooling as a covariate which was used as a surrogate for socio-economic status. Three categories of education level were defined: ‘low’ (up to 7 school years), ‘medium’ (8–10 school years) and ‘high’ (
11 school years).
In order to account for possible latency of the effect of radiation exposure on the development of glioma, ORs and 95% CIs were also calculated for a 10-year lag period prior to diagnosis. We tested for a linear trend across all the exposure categories in the logistic regression; trend P values were based on the Wald test.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
Four hundred and sixteen cases and 422 controls were recruited into the study accounting for 66% of the cases and 65% of the controls who were eligible and with whom contact was made. For the present analysis, two cases and one control had to be excluded due to the lack of occupational data. A total of 414 glioma cases (250 men and 164 women) and 421 controls (252 men and 169 women) were therefore included.
Occupation-specific exposure prevalences and levels of exposure for ELF, RF, UV and ionizing radiation as given by FINJEM for the period 1985–94 are shown in Tables 1 and 2. The tables illustrate the occupations with highest exposure as given by FINJEM and occupations with highest numbers of exposed cases.
|
|
Table 3 shows the ORs for the total cumulative exposure of the different radiation types as given by FINJEM. The majority of risk estimates were close to or below unity. The increased risks for RF (OR = 1.80, 95% CI 0.53–6.13) and UV (OR = 1.43, 95% CI 0.97–2.11) in the second tertile were statistically insignificant. The 10-year lagged data (not shown in Table 3) demonstrated no major difference from the total exposure. The ORs in the third tertile of exposure for ELF, RF, UV and ionizing radiation were 0.71 (95% CI 0.46–1.08), 0.58 (95% CI 0.14–2.47), 1.12 (95% CI 0.73–1.72) and 1.08 (95% CI 0.55–2.12), respectively.
|
The FINJEM gender analysis for UV is shown in Table 4. There was a positive trend for males and a negative trend for females, both of which were borderline statistically significant. Also from Table 4, it can be seen that the cumulative UV exposure was much higher for males than for females. The FINJEM analysis of ELF by gender did not show a significant association with glioma for either males or females (data not shown in Table 4). The ORs in the third tertile of the total cumulative ELF exposure for males and females were 0.70 (95% CI 0.42–1.15) and 1.12 (95% CI 0.59–2.14), respectively.
|
The risk estimates for RF and ionizing radiation, as given by self-reporting and expert assessment, are shown in Table 5. For RF, the ORs were all below unity when exposure was assessed by self-report. Elevated but statistically non-significant risk estimates were found when exposure was assessed by the expert hygienist. For ionizing radiation, the ORs for both self-report and expert assessment were generally close to or below unity. With a 10-year lag, the results for RF were similar to those for total exposure (data not shown in Table 5); the ORs in the third tertile of exposure were 0.51 (95% CI 0.19–1.39) and 1.66 (95% CI 0.61–4.50) for self-report and expert assessment, respectively. For ionizing radiation, the risk estimates were generally slightly lower with a 10-year lag (data not shown in Table 5); the ORs in the third tertile of exposure were 0.15 (95% CI 0.02–1.20) and 0.94 (95% CI 0.21–4.31) for self-report and expert assessment, respectively.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
Our results provide no support for an association between occupational exposure to ELF, RF or ionizing radiation and adult glioma. Although we observed occasional raised ORs, the CIs were wide and included unity. For UV radiation, we observed a borderline statistically significant elevated risk for men (P for trend = 0.03) but not for women. No previous epidemiological studies have specifically investigated a possible association between UV radiation exposure and glioma in adults.
During the past 20 years, a number of studies have examined the possibility that occupational exposure to ELF radiation increases the risk of brain tumour, although few have presented results for histologic subtypes. A meta-analysis of 29 early occupational studies reported a small increased risk for brain malignancy among electrical workers [14]. The exposure assessment in these early studies, however, has been criticized since it relied on job titles alone [15]. Several later industry-based studies performed on electric utility workers, which assessed individual exposures, have shown mixed results [16–20]. Our results are similar to a previous analysis of our data using a different JEM (designed in the United States) which was specifically designed to assess ELF radiation exposure [13] which produced an OR of 0.72 (95% CI 0.44–1.19) compared with our current estimate of 0.79 (0.53–1.18) for the highest exposed group of workers.
Results from studies investigating occupational RF radiation exposure and brain tumours have been mixed. Two reviews of epidemiological studies concluded that there was no consistent elevation of brain cancer risk associated with occupational RF exposure [21,22], which is consistent with our results. Our study improved on the exposure assessment used for earlier studies which relied on job titles alone, by assessing RF exposure using three methods: FINJEM, self-reports and expert hygienist review. It must be noted that exposure to low-level RF radiation such as mobile phone use was not considered; however, our study predates common use of mobile phones.
Although therapeutic ionizing radiation, which has higher exposure levels, is an established risk factor for brain tumours, most occupational studies have failed to find an association [23]. The ionizing radiation exposure levels in our study were low. A review by Inskip et al. [2] concluded that available data did not support the view that occupational exposure to ionizing radiation was causative of brain tumours and this is consistent with our results.
The strengths of our study need to be considered in comparison with many of the previous studies that have been mentioned. Our study was population based and one of the largest studies undertaken within the SEARCH program. The cases were all histologically confirmed, unlike many other death certificate-based occupational cohort and case–control studies. We only investigated one type of brain tumour, glioma, unlike the majority of previous studies which included a mix of histological subtypes, thus enhancing the likelihood of finding a real association (if one exists) for a particular type of radiation.
Another strength of our study was that FINJEM estimated exposure for different calendar time periods. Changes in exposure over time were, therefore, taken into account and relevant periods were used when considering the effect of latency. In addition, the specificity of FINJEM is relatively high since it provides estimates of both probability and level of exposure during the particular time periods [24]. In spite of this, misclassification of exposure is likely when exposure is estimated using surrogate measures such as job exposure matrices. Further, FINJEM's exposure estimates were developed in Finland and for certain radiation types (e.g. UV), the exposure for the same job titles may be quite different to that in Australia. The cumulative UV exposure estimates in Tables 2 and 3 are, therefore, probably underestimated because of this inter-country difference. For RF and ionizing radiation, exposure assessment included self-reports and expert hygienist review but due to lack of data this was not done for UV. In addition, large within-worker variations may have contributed to misclassification of exposure. Although, such non-differential misclassification generally tends to bias the risk estimates towards the null, the effect on our results is difficult to predict [25].
A limitation of our study was the use of proxies to complete the work history questionnaire for almost half of the cases but for only 2% of the controls. Although, proxies have poor knowledge about workplace exposures of subjects, in our study, they were asked only to provide information about job titles. While the results were unaltered when analyses were restricted to index subjects, the use of proxies was an important factor only for the cases, and any misclassification bias would be differential. The effect of this on the risk estimates is unknown.
The control subjects were chosen from the Electoral Roll which contained 85% of all people who were eligible to vote in 1989 [10]. The 15% not on the electoral roll had a strong non-English speaking bias but this is unlikely to be an important confounder. There were no other biases in those missing from the electoral roll. The Electoral Roll is updated prior to an election. Up to 10% of the population move address each year and the register of voters quickly gets out of date. There is no other population register available and the Electoral Roll is superior to the telephone directory as it contains the majority of male and female residents.
The study results are unlikely to be explained by confounding due to lifestyle factors. The subjects were matched by age and sex and although information was collected on cigarette smoking and alcohol consumption, these variables were not included in the final model of the analysis. Previous analyses of the same data [4] did not show an association between glioma and these two risk factors and this is consistent with the conclusions of several reviews [2,23,26].
In conclusion, only UV radiation was associated with increased glioma risk for men but not for women but this result could be a chance finding or due to other occupational and lifestyle exposures of men exposed occupationally to high levels of UV. Our results do not support an association between other radiation types and adult glioma. Given the paucity of information on UV exposure and glioma, we suggest that future studies should address the possibility of an association.
Key points
|
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
National Health and Medical Research Council of Australia, the Victorian Health Promotion Foundation and the Alfred Hospital.
|
Conflicts of interest |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
None declared.
|
Acknowledgements |
|---|
We thank all the subjects who participated; the nurse interviewers, J. Snaddon and S. Gardiner; L. Quango for computer programming assistance and the participating neurosurgeons.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingConflicts of interestReferences |
|---|
- Davis DL, Ahlbom A, Hoel D, Percy C. Is brain cancer mortality increasing in industrial countries? Am J Ind Med (1991) 19:421–431.[Web of Science][Medline]
- Inskip PD, Linet MS, Heineman EF. Etiology of brain tumours in adults. Epidemiol Rev (1995) 17:382–414.
[Free Full Text] - Giles GG, McNeil JJ, Donnan G, et al. Dietary factors and the risk of glioma in adults: results of a case-control study in Melbourne, Australia. Int J Cancer (1994) 59:357–362.[Web of Science][Medline]
- Hurley SF, McNeil JJ, Donnan GA, Forbes A, Salzberg M, Giles GG. Tobacco smoking and alcohol consumption as risk factors for glioma: a case-control study in Melbourne, Australia. J Epidemiol Community Health (1996) 50:442–446.
[Abstract/Free Full Text] - Cicuttini FM, Hurley SF, Forbes A, et al. Association of adult glioma with medical conditions, family and reproductive history. Int J Cancer (1997) 70:1–5.[CrossRef][Web of Science][Medline]
- Milham S. Mortality in workers exposed to electromagnetic fields. Environ Health Perspect (1985) 62:297–300.[Web of Science][Medline]
- Loomis DP, Savitz DA. Mortality from brain cancer and leukaemia among electrical workers. Br J Ind Med (1990) 47:633–638.[Web of Science][Medline]
- Grayson JK. Radiation exposure, socioeconomic status, and brain tumor risk in the US Air Force: a nested case-control study. Am J Epidemiol (1996) 143:480–486.
[Abstract/Free Full Text] - Health World Organization. In: International Classification of Diseases for Oncology (ICD-O) (1976) Geneva: WHO.
- Hurley SF. Screening: the need for a population register. Med J Aust (1990) 153:310–311.[Web of Science][Medline]
- Kauppinen T, Toikkanen J, Pukkala E. From cross-tabulations to multipurpose exposure information systems: a new job-exposure matrix. Am J Ind Med (1998) 33:409–417.[CrossRef][Web of Science][Medline]
- Benke G, Sim M, Fritschi L, Aldred G, Forbes A, Kauppinen T. Comparison of occupational exposure using three different methods: hygiene panel, job exposure matrix (JEM), and self reports. Appl Occup Environ Hyg (2001) 16:84–91.[CrossRef][Medline]
- Karipidis KK, Benke G, Sim MR, Yost M, Giles G. Occupational exposure to low frequency magnetic fields and the risk of low grade and high grade glioma. Cancer Causes Control (2007) 18:305–313.[CrossRef][Web of Science][Medline]
- Kheifets LI, Afifi AA, Buffler PA, Zhang ZW. Occupational electric and magnetic field exposure and brain cancer: a meta-analysis. J Occup Environ Med (1995) 37:1327–1341.[Web of Science][Medline]
- Kheifets LI. Electric and magnetic field exposure and brain cancer: a review. Bioelectromagnetics (2001) (Suppl. 5):S120–S131.
- Savitz DA, Loomis DP. Magnetic field exposure in relation to leukemia and brain cancer mortality among electric utility workers. Am J Epidemiol (1995) 141:123–134.
[Abstract/Free Full Text] - Miller AB, To T, Agnew DA, Wall C, Green LM. Leukemia following occupational exposure to 60-Hz electric and magnetic fields among Ontario electric utility workers. Am J Epidemiol (1996) 144:150–160.
[Abstract/Free Full Text] - Harrington JM, McBride DI, Sorahan T, Paddle GM, van Tongeren M. Occupational exposure to magnetic fields in relation to mortality from brain cancer among electricity generation and transmission workers. Occup Environ Med (1997) 54:7–13.
[Abstract/Free Full Text] - Johansen C, Olsen JH. Risk of cancer among Danish utility workers—a nationwide cohort study. Am J Epidemiol (1998) 147:548–555.
[Abstract/Free Full Text] - Savitz DA, Cai J, van Wijngaarden E, et al. Case-cohort analysis of brain cancer and leukemia in electric utility workers using a refined magnetic field job-exposure matrix. Am J Ind Med (2000) 38:417–425.[CrossRef][Web of Science][Medline]
- Elwood J. A critical review of epidemiologic studies of radiofrequency exposure and human cancers. Environ Health Perspect (1999) 107((Suppl.)):155–168.[Web of Science][Medline]
- Blettner M, Schlehofer B. Is there an increased risk of leukemia, brain tumors or breast cancer after exposure to high-frequency radiation? Review of methods and results of epidemiologic studies. Med Klin (1999) 15:150–158.
- Wrench M, Minn Y, Chew T, Bondy M, Berger MS. Epidemiology of primary brain tumors: current concepts and review of the literature. Neuro-oncol (2002) 4:278–299.[Abstract]
- Vasama-Neuvonen K, Pukkala E. Paakkulainen H et al. Ovarian cancer and occupational exposures in Finland. Am J Ind Med (1999) 36:83–89.[CrossRef][Web of Science][Medline]
- Copeland KT, Checkoway H, McMichael AJ, Holbrook RH. Bias due to misclassification in the estimation of relative risk. Am J Epidemiol (1977) 105:488–495.
[Abstract/Free Full Text] - Ohgaki H, Kleihues P. Epidemiology and etiology of gliomas. Acta Neuropathol (Berl) (2005) 109:93–108.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. Liu, M. E. Scheurer, R. El-Zein, Y. Cao, K.-A. Do, M. Gilbert, K. D. Aldape, Q. Wei, C. Etzel, and M. L. Bondy Association and Interactions between DNA Repair Gene Polymorphisms and Adult Glioma Cancer Epidemiol. Biomarkers Prev., January 1, 2009; 18(1): 204 - 214. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||