Content area
Full Text
Introduction
Inorganic arsenic is a naturally occurring set of compounds that is classified as a known human carcinogen and is of potential concern because of its ubiquity in soil and water and the systemic nature by which it influences disease in humans (IARC et al. 2004; U.S. EPA 2001). Evidence regarding the association between ingested arsenic and human health has come mainly from populations exposed to high levels of arsenic in drinking water. In those studies, arsenic has been associated with increases in all-cause mortality [e.g., Argos et al. (2010)], ischemic heart disease [e.g., Tseng et al. (2003)], nonmalignant respiratory disease [e.g., Mazumder et al. (2000)], and cancers at sites including lung, skin, bladder, and kidney [e.g., Wu et al. (1989)]. In contrast, most epidemiological evidence regarding the health effects of inhaled inorganic arsenic comes from studies of exposure in workplaces. In such settings, the most consistent evidence of association with exposure to arsenic has been for lung cancer [e.g., Lubin et al. (2000)]. There is disagreement about whether data on airborne arsenic exposure provide evidence of associations with other health outcomes [e.g., Hertz-Picciotto et al. (2000); Lubin and Fraumeni (2000)]. Lung cancer is relatively rare. Thus, the population impact of airborne arsenic exposure has been described mainly by high relative rates for a rare disease, which limits the apparent magnitude of the potential public health impact.
It is not clear whether the lack of associations between airborne arsenic exposure and other causes of death is due to differences between the health effects by exposure route or whether it is due to bias that obscures health effect estimates in occupational studies. Previous authors have speculated that such associations may be masked by healthy worker survivor bias (Arrighi and Hertz-Picciotto 1994), a bias in occupational studies that may obscure causal relationships. Healthy worker survivor bias occurs when healthier workers (those experiencing low rates of disease) are able to sustain employment longer than sicker workers, thus accruing greater exposure over time. Some authors assert that healthy worker survivor bias does not appreciably affect estimates of the health effects of arsenic (Lubin and Fraumeni 2000). We propose that progress in assessing the magnitude of the public health impact of inhaled arsenic can be made by reanalyzing occupational data...