br Secondary analysis br If analyses
3.4. Secondary analysis
If analyses were extended to exposure/cancer site pairs where there is limited evidence of a causal relationship, then 12,314 new cancer cases, representing 3.5% of all new cancer cases, were estimated to be
attributable to occupational agents in France in 2015 (Table S3 of Supplemental material). As also observed in the main analysis, the occupational exposures disproportionately aﬀected men, with 10,814 new cancer cases among men, representing 5.7% of all new cancer cases among men, and 1500 new cancer cases among women, representing 1.0% of all new cancer cases among women. The inclusion of two ex-posures/cancer site pairs associated with limited evidence of a causal association in humans had a noticeable impact on the number of at-tributable cases: exposure to asbestos added 758 attributable colorectal cancer cases in men and 70 in women; and exposure to shift work in women added 669 breast cancer cases (Table S4 of Supplemental ma-terial).
This study presents the first, most comprehensive, estimation for France of the number of new cancer cases attributable to occupational exposures that accounts for the FSL1 of the exposure over time. Specifically, this study estimated that 7900 new cancer cases, re-presenting approximately 2.3% of all new cancer cases (3.9% among males and 0.4% among women), were attributable to occupational exposures. When extending the analysis to exposure/cancer site pairs
Fig. 1. Agents' contribution to the total number of cancer cases attributable to occupational exposures by sex in France in 2015 (including only agents and cancer sites with suﬃcient evidence of a causal association in humans).
C. Marant Micallef et al. International Journal of Hygiene and Environmental Health 222 (2019) 22–29
Estimated number of cases (N) and fraction of cancers cases (PAF) attributable to occupational exposures, by exposure.a
Agents (Group 1) Cancer sites (ICD-10) Men
a These results are for carcinogenic agents (IARC Monograph Group 1) where there is suﬃcient evidence of a causal relationship between exposure and the development of cancer in humans.
with the limited level of evidence of a causal relationship, an estimated 12,300 new cancer cases, representing 3.5% of all new cancer cases (5.7% among males and 1.0% among women), were attributable to occupational exposures.
We found that lung cancers were by far the leading type of cancer among men and women attributable to occupational exposure. Among the main agents related to lung cancers in France, some are still present in the working environment (Chromium VI, silica, diesel engine ex-haust) (Vinck and Memmi, 2015). Many occupational carcinogens in-cluding Chromium VI, silica or diesel engine exhaust have been sug-gested to have common exposure route through the respiratory system (Delva et al., 2016; Field and Withers, 2012). Therefore, monitoring of air contaminant concentrations and reduction of exposure in various working environments (Institut National Du Cancer, 2013; World Health Organization, 2013) could prevent a large number of new lung cancer cases and lead to a decrease in occupational related cancers.
4.1. Comparison to other studies
The percentage of all new cancer cases attributable to occupational exposures estimated by this study was higher than that estimated by a previous study for France in 2000, where 1.6% of all new cancer cases were attributable to occupational exposures (2.7% among males and 0.3% among women) (Boﬀetta et al., 2010). In both studies the largest numbers of new cancer cases were observed for the same cancer sites (i.e., lung cancer in both males and females, followed by mesothelioma and bladder cancers in men). The current study estimated the number
of new cancer cases attributable to a wider range of occupational ex-posures (and also additional cancer sites associated with these ex-posures), which might explain part of the diﬀerences in the estimates. Diﬀerences in data sources used, as well as the method to estimate the prevalence of exposure may be other sources of discrepancies. Fur-thermore, the findings of the current study were similar to the recent estimates of the French National Public Health Agency for 2012 (for the four common agents reported in both studies) (Gilg Soit Ilg et al., 2016).
In other countries the percentage of new cancer cases attributable to occupational agents have ranged from 3% to 14% among men and from 0% to 2% of all new cancer cases among women (Boﬀetta et al., 2010; Driscoll et al., 2005; Fritschi and Driscoll, 2006; Nurminen and Karjalainen, 2001; Purdue et al., 2015; Rushton et al., 2010). However, direct comparisons between this study and other previous studies are complicated due to diﬀerences in the occupational agents and agent/ cancer site pairs included, the RR estimates and the methodologies used (Purdue et al., 2015), as well as in the exposure prevalences. A study in the UK in 2010 used a similar methodology to this current study and resulted in similar estimates. Specifically, in the UK in 2010, 4.0% of all new cancer cases, 5.7% among men, and 2.2% among women, were attributable to occupational agents (including pairs with suﬃcient and limited evidence of a causal association with cancer in humans). Ad-ditionally, similar major contributors to the attributable number of cases (asbestos, silica, work as a painter in men, shift work in women) were observed between the current study and those estimated for the UK in 2010 (Rushton et al., 2010).