INTEROCC case–control study: lack of association between glioma tumors and occupational exposure to selected combustion products, dusts and other chemical agents
1 University of Montreal Hospital Research Centre (CRCHUM), Montreal, Canada
2 Center for Research on Environmental Epidemiology (CREAL), Barcelona, Spain
3 Oregon State University, Corvallis, USA
4 Monash University, Melbourne, Australia
5 University of Leeds, Leeds, UK
6 INRETS, Lyon, France
7 University of Ottawa, Ottawa, Canada
8 Massey University, Wellington, New Zealand
9 INRS-Institut Armand-Frappier|, Montreal, Canada
10 Gertner Institute, Chaim Sheba Medical Center & Tel Aviv University, Tel Aviv, Israel
11 DFKZ, Heidelberg, Germany
12 Centre for Human Exposure Science, Institute of Occupational Medicine, Edinburgh, UK
BMC Public Health 2013, 13:340 doi:10.1186/1471-2458-13-340Published: 12 April 2013
The aim was to investigate possible associations between glioma (an aggressive type of brain cancer) and occupational exposure to selected agents: combustion products (diesel and gasoline exhaust emissions, benzo(a)pyrene), dusts (animal dust, asbestos, crystalline silica, wood dust) and some other chemical agents (formaldehyde, oil mist, sulphur dioxide).
The INTEROCC study included cases diagnosed with glioma during 2000–2004 in sub-regions of seven countries. Population controls, selected from various sampling frames in different centers, were frequency or individually matched to cases by sex, age and center. Face-to-face interviews with the subject or a proxy respondent were conducted by trained interviewers. Detailed information was collected on socio-economic and lifestyle characteristics, medical history and work history. Occupational exposure to the 10 selected agents was assessed by a job exposure matrix (JEM) which provides estimates of the probability and level of exposure for different occupations. Using a 25% probability of exposure in a given occupation in the JEM as the threshold for considering a worker exposed, the lifetime prevalence of exposure varied from about 1% to about 15% for the different agents. Associations between glioma and each of the 10 agents were estimated by conditional logistic regression, and using three separate exposure indices: i) ever vs. never; ii) lifetime cumulative exposure; iii) total duration of exposure.
The study sample consisted of 1,800 glioma cases and 5,160 controls. Most odds ratio estimates were close to the null value. None of the ten agents displayed a significantly increased odds ratio nor any indication of dose–response relationships with cumulative exposure or with duration of exposure.
Thus, there was no evidence that these exposures influence risk of glioma.