br Fig Standard mortality ratios by possible occupational ex
Fig. 1. Standard mortality ratios by possible occupational exposure and no known occupational exposure.
Mesothelioma standardized mortality ratios by possible occupational exposure and gender.
No occupational exposure Possible occupational exposureb
SMR, Standard mortality ratios.
*Statistically significant. aFollow-up period 2001–2011.
b Work in a profession where asbestos exposure is common (electrician, mechanic, brake repair, welding, ship building, home construction, ship building or repair, carpentry, mining, plumbing, pipefitting, roofing, insulation work, installing/removing flooring, other work with asbestos, worked at ver-miculite processing plant).
identified two cases in females with no direct occupational or take-home exposure; their only known exposure was from ambient air ex-posure as residents for over 20 years. This supports a hypothesis that cumulative, low-level environmental exposure to Libby asbestos through multiple pathways may have caused mesothelioma. Our finding of elevated COPD mortality among females is also consistent with the recent 33-year mortality study in the Libby, Montana popu-lation (Naik et al., 2017). Unlike the Libby analysis we did not observe a significant excess of asbestosis deaths.
There are several key strengths to this analysis. The Minneapolis processing site had an extensive 50-year history of operating near a densely populated, urban residential community. The discovery of widespread distribution of amphibole waste rock that contaminated over 260 residential properties was unique and instigated the detailed characterization of multiple pathways of exposure from over 6400 NMCVI interviews, records searches, and ambient air emissions mod-eling, which have been documented in previous publications. The as-sessment recorded detailed residential history for all addresses in the study area, history of occupations with possible asbestos exposure, and smoking histories for all participants. The NMCVI exposure assessment demonstrated that many of the non-occupational asbestos exposure pathways in this Minneapolis PD 98059 were similar to those identi-fied in Libby. Furthermore, the statewide cancer surveillance system used for data linkage has been operating since 1988 and serves as the population-based registry, using hospital-based and microscopically confirmed cancers for near complete case-finding. We conducted sensitivity analyses to assess the impact of occupa-tional exposure or pile playing, an activity with a high intermittent level of exposure. To assess whether the elevated mortality risk
Standardized mortality ratios by log cumulative ambient asbestos exposure levels.
estimates may be attributed to the inclusion of WM/WRG workers in the study populations, we removed the workers (n = 26) and repeated the analysis. We determined that the mesothelioma and lung cancer SMRs were still significant when the population was restricted to only community members with no occupational exposure. The history of children from the neighborhood playing on the waste rock piles was a reported activity by 11% of the 5848 study participants in this analysis. To assess whether the elevated SMRs were attributed to pile players, a group with high intermittent asbestos exposure, the pile players (n = 34) were removed from the analysis; the mesothelioma and lung cancer SMRs were still significantly elevated. A history of smoking is a significant risk factor that may confound these findings of increased lung cancer in both genders. Smoking was highly prevalent (57%) in the study population and is estimated to account for 80–90% of all lung cancers. Compared to state smoking levels reported in the 2001 Minnesota Behavioral Risk Factor Surveillance System, the percent of study participants who reported ever smoking was 7% greater in men and 6.6% greater in women. Therefore, the excess of lung cancer observed in these results may be largely due to an excess of smoking and not asbestos exposure. Additionally, in a separate analysis, we observed no dose-response pattern with increasing ambient asbestos exposure and lung cancer risk among smokers, when stratified by pack-year categories. However, we cannot rule out the possibility that lung cancer rates may be elevated due to the known synergistic effects of smoking and asbestos exposure (Ngamwong et al., 2015).
A lower incidence of other cancers can lead to inflation of the PIR for the cancer of interest. We estimated the SIR and overall cancer deficit in the community and adjusted for this to demonstrate that the PIRs remained significant. To assess for the possible impact of out-mi-gration from the state, we examined the total number of new deaths captured through the NDI search compared to the number found in Minnesota vital records as an indicator of the extent cohort participants may have moved from the area. Of the 847 deaths, there were 21 out-of-state deaths, indicating that the majority (98%) of the deceased co-hort stayed in Minnesota.