High concentrations of arsenic in drinking water result in the highest known increases in mortality attributable to any environmental exposure Allan H. Smith Professor of Epidemiology University of California, Berkeley Arsenic Health Effects Research Group Presented August 29, 2007 Royal Geographic Society Annual International Conference, London
Co-authors Craig Steinmaus, Yan Yuan, Jane Liaw, Meera M Hira-Smith
Some of our key collaborators: Chile, Dr. Catterina Ferreccio Dr. Guillermo Marshall India, Dr. Guha Mazumder Dr. Ashok Giri Bangladesh, Dr Md.Yunus Argentina, Dr. Omar Rey University of Washington, Dr. David Kalman UCSF (San Francisco), Dr. John Balmes UC Berkeley, Dr. Martyn Smith Dr. Alex Arroyo
The idea that something in drinking water might have major impacts in the lung is: A little? incredible
Could invisible tasteless and odorless arsenic in drinking water really affect the lungs?
Known causes of lung cancer involve inhalation smoking passive smoking asbestos radon silica chromium diesel exhaust coke oven PAHs bischlormethyl ether nickel arsenic
Arsenic in drinking water It is ingested, not inhaled Could it possibly affect the lungs and other organs like cigarette smoking? And could the risks be compared to those from smoking cigarettes? Incredible!
Map of Chile http://www.ine.cl/territor iales/chile.jpg Region II Region V CHILE
Arsenic concentrations in Antofagasta and Mejillones water by year. Arsenic contaminated water sources were used from 1958, and an arsenic removal plant was installed in 1971. 900 800 700 600 500 400 300 200 100 0 1950-57 1958-70 1971-80 1981-90 1991-2000
Mortality from lung cancer among men, Region II Chile Marshall et al, J. Natl Cancer Inst, 2007 5 Peak exposure 4 started stopped Rate Ratios 3 2 1 0 1950 1960 1970 1980 1990 2000 Year Rate ratios Lower 95% CI Upper 95% CI
Relative Risk Rela tive Risk 18 16 14 12 10 8 6 4 2 0 18 16 14 12 10 8 6 4 2 0 Lung Cancer Relative Risk Estimates Lung Cancer Relative Risk Estimates Taiwan (males) Argentina Tai wan (males) Argentina (males) Chile (male) Chile (male) Chile (case-control) Chile (case-control) s Taiwan (females) Argentina (females) Taiwan (females) Argentina (females) Chile (female) Japan (cohort) 0 400 800 1200 1600 s Chile (female) Japan (cohort) 0 Arsenic 400 Concentrations 800 1200 (ug/l) 1600 Arsenic Concentrations (ug/l)
IARC classification, 2002 Arsenic in drinking-water was evaluated as carcinogenic to humans (Group 1) on the basis of sufficient evidence for an increased risk for cancer of the urinary bladder, lung and skin.
Childhood exposure
Lung cancer mortality in men according to exposure in childhood born 1989 1998 peak arsenic 1950 1960 1970 1980 1990 2000
Lung cancer mortality in older men (SMR = standardized mortality ratio = observed/expected deaths) born 1920-38 Age at death 60 69 1989 1998 14 12 10 8 6 SMR peak arsenic 1950 1960 1970 1980 1990 2000 4 2 0 rest Chile exposed p < 0.001
Lung cancer mortality in older men (SMR = standardized mortality ratio = observed/expected deaths) born 1943-47 Age at death 45 49 1989 1998 14 12 10 8 6 SMR peak arsenic 1950 1960 1970 1980 1990 2000 4 2 0 rest Chile exposed p < 0.001
Lung cancer mortality in men according to exposure in childhood (SMR = standardized mortality ratio = observed/expected deaths) born 1953-57 Age at death 35 39 1989 1998 14 12 10 8 6 SMR peak arsenic 1950 1960 1970 1980 1990 2000 4 2 0 rest Chile exposed p < 0.001
Lung cancer mortality in men according to exposure in childhood (SMR = standardized mortality ratio = observed/expected deaths) born after 1957 peak arsenic Age at death 30 34 1989 1998 1950 1960 1970 1980 1990 2000 14 12 10 8 6 4 2 0 SMR rest Chile exposed p < 0.001
9 8 7 6 5 4 3 2 1 0 Childhood exposures and young adult lung cancer relative risks atom bomb ETS arsenic Smith AH et al. Environmental Health Perspectives 2006 Atomic bomb survivors Hiroshima and Nagasaki aged 10-20 in 1945. No cases in those exposed in utero or under age 10. A study giving lung cancer relative risks in the age range 30-39 following many hours of daily passive smoking as children (only four cases). Lung cancer relative risks in age range 30-39 in Region II of Chile following early life exposure to arsenic in water. There were 32 lung cancer deaths with 3.8 expected (RR=8.4, p<0.001).
Arsenic in drinking water results in the highest known adult cancer risks from any childhood exposure
Overall respiratory study design Initial survey in 1995/96 of 7,600 people 415 were found to have arsenic skin lesions We selected persons with skin lesions living in 21 villages whose primary drinking water sources contained up to 500 g/l.
Lung function findings of reduced FEV1 adjusted for age and height 0-50 For all men combined P=0.007-100 -150-200 -250-300 men skin lesions men, smoking women, skin lesions Among men in this population, arsenic- caused skin lesions were associated with a greater FEV1 reduction (-256ml) than from smoking (-156ml) Von Ehrenstein OS et al. Am J Epidemiol 162: 533-41, 2005
High resolution computed tomography (HRCT) with readings in India and the United States without knowing who had skin lesions.
Study design, x-ray (HRCT) study in West Bengal, India 7600 surveyed Assessed in this study 108 selected with skin lesions and high arsenic exposure 150 selected with no skin lesions and low arsenic exposure Referred for HRCT 27 had chronic cough more than 2 years 11 had chronic cough More than 2 years 18 had evidence of bronchiectasis 3 had evidence of bronchiectasis 10-fold increased prevalence of bronchiectasis OR=10.1, p<0.01
End Stage Lung Disease Obstruction, Infection, Hemoptysis, Bronchiectasis
Source: Project Well, West Bengal, India, 2003
Arsenic in Drinking Water and Skin Lesions: Dose-Response Data from West Bengal, India Haque R et al. Epidemiology 2003 The lowest peak arsenic water concentration ingested by a confirmed case was 115 ug/liter. All other cases had peak water concentrations of at least 150 ug/l Note: Detailed exposure assessment is essential
Arsenic concentrations in Antofagasta and Mejillones water by year. Arsenic contaminated water sources were used from 1958, and an arsenic removal plant was installed in 1971. 900 800 700 600 500 400 300 200 100 0 1950-57 1958-70 1971-80 1981-90 1991-2000
Chronic obstructive pulmonary disease (COPD) mortality in young men according to exposure in childhood born 1989 2000 peak arsenic 1950 1960 1970 1980 1990 2000
Chronic obstructive pulmonary disease (COPD) mortality in young men according to exposure in childhood born 1989 2000 peak arsenic 1950 1960 1970 1980 1990 2000
Mortality (SMRs) from Chronic Obstructive Pulmonary Disease, age 30-49, for those born in the very high exposure period (in utero) or just before (child) 16 14 12 10 8 6 4 2 0 All COPD All in utero child P<0.001
Mortality (SMRs) from Chronic Obstructive Pulmonary Disease, age 30-49, for those born in the very high exposure period (in utero) or just before (child) 50 45 40 35 30 25 20 15 10 5 0 All COPD All in utero child
Mortality (SMRs) from Chronic Obstructive Pulmonary Disease, age 30-49, for those born in the very high exposure period (in utero exposure) or just before (child) 50 45 40 35 30 25 20 15 10 5 0 All COPD All in utero child p<0.001 except other COPD p=0.004 50 45 40 35 30 25 20 15 10 5 0 50 45 40 35 30 25 20 15 10 5 0 Bronchiectasis Other COPD All in utero child All in utero child
Map of Chile http://www.ine.cl/territor iales/chile.jpg Region II Region V CHILE
Male mortality from acute myocardial infarction Peak exposure started stopped 2.0 Male 1.5 Rate ratios 1.0 0.5 0.0 1950 1960 1970 1980 1990 2000 Yuan Y et al. Am J Epidemiol. In Press
Female mortality from acute myocardial infarction Peak exposure started stopped 2.0 Female 1.5 Rate ratios 1.0 0.5 0.0 1950 1960 1970 1980 1990 2000 Year Yuan Y et al. Am J Epidemiol. In Press
Excess deaths attributable to arsenic in Region II Chile Years Total Deaths Excess Deaths Due to Acute Myocardial Infarction Excess Deaths Due to Lung Cancer Excess Deaths Due to Bladder Cancer Total Excess Deaths Excess Deaths as a Percentage of Total Deaths (%) Male: 1950-1957 1958-1964 1965-1970 1971-1979* 1980-1985 1986-1990 1991-1995 1996-2000 Total Female: 1950-1957 1958-1964 1965-1970 1971-1979* 1980-1985 1986-1990 1991-1995 1996-2000 Total 5,604 5,650 5,025 7,966 6,285 5,152 5,639 5,944 47,265 3,722 3,596 3,251 5,158 3,998 3,793 4,079 4,568 32,165 29 222 230 267 195 154 115 40 1,252 9 45 84 137 12 8 5 0 300 17 9 62 229 303 305 412 358 1,695 4 7 12 41 39 84 92 113 392 10 6 11 20 64 60 86 73 330 5 5 5 17 39 50 65 69 255 56 237 303 516 562 519 613 471 3,277 18 57 101 195 90 142 162 182 947 1.00 4.19 6.03 6.48 8.94 10.07 10.87 7.92 6.93 0.48 1.59 3.11 3.78 2.75 3.85 4.00 3.36 2.94
Acute myocardial infarction, lung cancer and bladder cancer mortality after exposure to arsenic Acute Myocardial Infarction Lung Cancer Bladder Cancer 12 Male 10 Excess deaths (%) 8 6 4 2 0 1950-1957 1958-1964 1965-1970 1971-1979 1980-1985 1986-1990 1991-1995 1996-2000
Acute myocardial infarction, lung cancer and bladder cancer mortality after exposure to arsenic 6 Female Excess deaths (%) 4 2 0 1950-1957 1958-1964 1965-1970 1971-1979 1980-1985 1986-1990 1991-1995 1996-2000 Years
The estimated cancer risk at the drinking water standard of 50 µg/l for arsenic is more than 100 times greater than that for any other drinking water contaminant Smith AH, Lopipero PA, Bates MN, Steinmaus CM. Arsenic epidemiology and drinking water standards. Science 296: 2145-6, 2002
Cancer risk from contaminants in drinking water other than arsenic Per 100,000 14 12 10 Top of the list: Ethylene dibromide 8 6 4 2 0
Cancer risk from contaminants in drinking water including arsenic Per 100,000 1400 1200 1000 800 Cancer risks from arsenic at the old drinking water standard were >100 times higher than the next highest risk contaminant 600 400 200 0
Contamination of drinking-water by arsenic in Bangladesh: a public health emergency Allan H. Smith et al The fundamental intervention is the identification and provision of arsenic-free drinking water. Arsenic is rapidly excreted in urine, and for early or mild cases, no specific treatment is required. Community education and participation are essential to ensure that interventions are successful; these should be coupled with follow-up monitoring to confirm that exposure has ended. Bulletin of the World Health Organization, 2000, 78: 1093-1103
Cancer risks from arsenic in drinking water Lung cancer and smoking 10 ug/l 50 ug/l 500 ug/l 5000 ug/l 1 in 500 die 1 in 100 die 1 in 10 die all die married to a smoker active smoker And arsenic in water looks good, does not smell and has no taste. So the risks are incredible
High concentrations of arsenic in drinking water result in the highest known increases in mortality attributable to any environmental exposure All wells in the world should be tested for arsenic Fortunately, arsenic is excreted rapidly and no treatment is required after stopping exposure The end