Start | Committee Index | Author Index

226 Drinking Water Arsenic and Lung Cancer Risk – Non-Linearity Across the Literature

Wednesday, June 17, 2015: 10:00 AM-10:30 AM
Exhibit Hall A, Hynes Convention Center
Hamid Ferdosi , Milken Institute School of Public Health, The George Washington University, washington, DC
Nana Ama Afari-Dwamena , Consultants in Epidemiology and Occupational Health, Washington, DC
Elisabeth Dissen , Consultants in Epidemiology and Occupational Health, washington, DC
Ji Li , Johns Hopkins University School of Medicine, Baltimore, MD
Lu Qian , American University, washington, DC
Manning Feinleib , Consultants in Epidemiology and Occupational Health, washington, DC
Steven H. Lamm , Center for Epidemiology and Global Health, washington, DC

BACKGROUND: High level arsenic levels (>200 ug/L) are associated with increased lung cancer risk; Low level arsenic levels (< 100 ug/L) less certainly and possibly flat or negative.  A systematic review has been conducted in preparation for a meta-analysis.      

METHODS: Comprehensive electronic literature search yielded a final set of 11 papers with 17 study populations that reported the risks of lung cancer from multiple levels of arsenic exposure in drinking water.  The risk pattern for individual studies and all studies were demonstrated in a plot of log relative risk versus log arsenic over a range of 1 ug/L to nearly1,000 ug/L. 

RESULTS: Risk estimates were generally in the range of 0.75-1.15 for exposure levels of 3 ug/L to 75 ug/L  Twenty-one risk estimates at exposures below 100 ug/L had an average risk of 0.96, an interquartile range of 0.75 to 1.14, and a range of 0.57 to 1.37.  Fifteen risk estimates at exposures off 100 ug/L to 400 ug/L had an average risk of 1.47, an interquartile range of 0.94-1.74, and a range of 0.84 to 3.40. Exposures in the 400 to 600 ug/L were 2.0 and 2.3, and 10 exposures in the 600-900 ug/L range had an average of risk of 3.7, an interquartile range of 3.0 to 4.7, and a range of 1.57 to 7.0. An analysis of 48 risk estimates from 17 study populations in 11 studies showed no increased risk below 100 ug/L arsenic with progressively increased risks from100 ug/L to1,000 ug/L arsenic.

CONCLUSIONS: Increased risks of lung cancers began to be observed at 100-200 ug/L arsenic and at >200 ug/L arsenic, but were not seen at < 100 ug/L arsenic.  The best fitting models across all study designs was a linear-quadratic fit with a significant negative coefficient to the linear term and may suggest no evidence of an increase in lung cancer risk generally at exposures below about 75 ug/L arsenic based on the non-ecological studies.

See more of: Environmental Health
See more of: Poster Sessions
<< Previous Abstract | Next Abstract >>