有色金属冶炼厂周围居民镉暴露致癌风险评估

目的了解有色金属冶炼厂周围镉污染现状,评估调查区居民镉暴露致癌风险。方法对湖北某有色金属冶炼厂周围地区多种环境介质[土壤、地下水、大气及农作物((蔬菜和大米)]进行采样,分析其中镉(Cd)含量,利用美国EPA致癌风险评估模型评估研究调查区居民经多种途径接触多种环境介质引起的Cd的致癌暴露剂量和致癌风险。结果调查区土壤、大气及农作物均存在镉污染,蔬菜和大米中镉的平均含量分别为我国食品中污染物限量标准的14倍和5倍,土壤中镉均不同程度超标,大气中镉含量均高于国内外其他地区的调查结果。镉经所有暴露途径总致癌风险高达2.2×10~(-3),远高于国内外所规定的可接受致癌风险水平;致癌贡献率分析结果表明蔬菜和大米是该地区最主要的镉致癌风险来源。结论有色金属冶炼厂周围调查区域环境受到了不同程度的镉污染,且对居民的健康存在风险。     

Leachate Properties and Cadmium Migration Through Freeze-thaw Treated Soil Columns

Soil column leaching experiments were conducted to study the effects of multiple freeze-thaw cycles on the vertical migration of cadmium (Cd). Three Cd-spiked leaching solutions of different properties were derived from snowmelt, sludge, and straw, designated as B, W and J, respectively. The leaching solutions varied in dissolved organic matter (DOM) concentrations in the order of J > W > B. Changes in leachate properties and Cd concentration were observed. The results showed that pH values of all the leachate solutions through freeze-thaw treated soil columns were higher than those of leachates through unfrozen soils. However, electrical conductivity (EC) values decreased compared with leachates in unfrozen treated soil columns. Although the concentrations of DOM in leachate solutions had no evident differences between the freeze-thaw and unfrozen treated soil columns, the concentrations of DOM in the leachate solutions B, W and J were different. Freeze-thaw cycles resulted in increased concentrations of Cd in the leachate solutions in the order J > W > B, and promoted a deeper migration of Cd in the soil columns. Thus, it was shown that freeze-thaw cycles may increase the risk of groundwater pollution by Cd.     

A GIS-aided assessment of the health hazards of cadmium in farm soils in central Taiwan

A geostatistical method was developed to examine the correlation, or lack of it, between the levels of cadmium (Cd) detected in farm soils and those detected in the human specimens collected from residents around the contaminated areas in Changhua County where cadmium contamination of staple rice has been documented. We used the Taiwan EPA environment data in 2002 and human data which were generated by the National Health Research Institutes during 2003-2005. Kriging interpolation methods were used to determine soil Cd concentrations. A Zonal statistical function was performed to assess the individual exposure. Soil Cd levels and tissue Cd levels in residents were analyzed for contamination hotspots and other areas to determine correlation between the two variables. Three Cd contamination hotspots were identified, in which no correlation was found between soil Cd levels and tissue Cd levels in residents. Our results demonstrate how GIS spatial modeling technique can be used to estimate distribution of pollutants in an area using a limited number of data points. Results indicated no association between the soil contamination and the exposure of residents to Cd, suggesting that both the soils and the residents are receptors of Cd as a pollutant from as yet unidentified sources.     

Personal exposure meets risk assessment: a comparison of measured and modeled exposures and risks in an urban community

Human exposure research has consistently shown that, for most volatile organic compounds (VOCs), personal exposures are vastly different from outdoor air concentrations. Therefore, risk estimates based on ambient measurements may over- or underestimate risk, leading to ineffective or inefficient management strategies. In the present study we examine the extent of exposure misclassification and its impact on risk for exposure estimated by the U.S. Environmental Protection Agency (U.S. EPA) Assessment System for Population Exposure Nationwide (ASPEN) model relative to monitoring results from a community-based exposure assessment conducted in Baltimore, Maryland (USA). This study is the first direct comparison of the ASPEN model (as used by the U.S. EPA for the Cumulative Exposure Project and subsequently the National-Scale Air Toxics Assessment) and human exposure data to estimate health risks. A random sampling strategy was used to recruit 33 nonsmoking adult community residents. Passive air sampling badges were used to assess 3-day time-weighted-average personal exposure as well as outdoor and indoor residential concentrations of VOCs for each study participant. In general, personal exposures were greater than indoor VOC concentrations, which were greater than outdoor VOC concentrations. Public health risks due to actual personal exposures were estimated. In comparing measured personal exposures and indoor and outdoor VOC concentrations with ASPEN model estimates for ambient concentrations, our data suggest that ASPEN was reasonably accurate as a surrogate for personal exposures (measured exposures of community residents) for VOCs emitted primarily from mobile sources or VOCs that occur as global "background" source pollutant with no indoor source contributions. Otherwise, the ASPEN model estimates were generally lower than measured personal exposures and the estimated health risks. ASPEN's lower exposures resulted in proportional underestimation of cumulative cancer risk when pollutant exposures were combined to estimate cumulative risk. Median cumulative lifetime cancer risk based on personal exposures was 3-fold greater than estimates based on ASPEN-modeled concentrations. These findings demonstrate the significance of indoor exposure sources and the importance of indoor and/or personal monitoring for accurate assessment of risk. Environmental health policies may not be sufficient in reducing exposures and risks if they are based solely on modeled ambient VOC concentrations. Results from our study underscore the need for a coordinated multimedia approach to exposure assessment for setting public health policy.     

贵州某环境污染区人群镉暴露水平分析

目的 了解贵州某镉污染区农作物污染状况及人群镉暴露水平.方法 通过随机抽样方法获得调查人群(1 349人),采用三日询问法获得人群膳食消费量,原子吸收法测定膳食镉含量,电感耦合等离子质谱检测尿镉用于内暴露评估.结果 采集粮食样品957份.完成186人的膳食调查和1 349人尿样筛查.污染区玉米、土豆及青菜的镉含量超过食...     

Modeling Cadmium Exposures in Low- and High-Exposure Areas in Thailand

Background: Previous U.S. population modeling studies have reported that urinary cadmium (Cd) excretion patterns differ with age, sex, and dietary exposure; associations between Cd exposures and health outcomes also have differed by age and sex. Therefore, it is important to test models used to estimate Cd exposures across an expanded Cd-exposure range.Objectives: We estimated relative Cd exposures from both diet and smoking in low- and high-exposure scenarios to provide data for improving risk assessment calculations.Methods: We used a Cd toxicokinetic–based model to estimate Cd exposures based on urinary Cd levels measured for 399 persons in a low-exposure area (Bangkok) and 6,747 persons in a high-exposure area (Mae Sot) in Thailand.Results: In Bangkok, we estimated dietary Cd exposures of 50–56 µg/day for males and 21–27 µg/day for females 20–59 years of age who never smoked. In Mae Sot, we estimated dietary Cd exposures of 188–224 µg/day for males and 99–113 µg/day for females 20–59 years of age who never smoked. In Bangkok, we estimated Cd exposures from smoking to be 5.5–20.4 µg/day for male smokers 20–59 years of age. In Mae Sot, we estimated Cd exposures from smoking to be 9.8–26 µg/day for male heavy smokers and 26 µg/day for female heavy smokers.Conclusion: This study provides estimates of Cd exposures from diet and smoking in low- and high-exposure scenarios. Our findings suggest a relatively small safety margin between the established tolerable Cd reference exposure of 62 µg/day and exposure levels previously associated with evidence of kidney and bone effects in Mae Sot residents, where dietary Cd exposures among women were only 1.6–2.1 times the reference value.     

Metal Exposures at three U.S. electronic scrap recycling facilities

Many metals found in electronic scrap are known to cause serious health effects, including but not limited to cancer and respiratory, neurologic, renal, and reproductive damage. The National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention performed three health hazard evaluations at electronic scrap recycling facilities in the U.S. to characterize employee exposure to metals and recommend control strategies to reduce these exposures. We performed air, surface, and biological monitoring for metals. We found one overexposure to lead and two overexposures to cadmium. We found metals on non-production surfaces, and the skin and clothing of workers before they left work in all of the facilities. We also found some elevated blood lead levels (above 10 micrograms per deciliter), however no employees at any facility had detectable mercury in their urine or exceeded 34% of the OELs for blood or urine cadmium. This article focuses on sampling results for lead, cadmium, mercury, and indium. We provided recommendations for improving local exhaust ventilation, reducing the recirculation of potentially contaminated air, using respirators until exposures are controlled, and reducing the migration of contaminants from production to non-production areas. We also recommended ways for employees to prevent taking home metal dust by using work uniforms laundered on-site, storing personal and work items in separate lockers, and using washing facilities equipped with lead-removing cleaning products.     

Cadmium exposure pathways in the Czech urban population

The article describes the exposure pathways of cadmium in the Czech urban population. The data on Cd concentrations originated from the Environmental Health Monitoring System, which has been realized in 30 cities since 1994. The data on cadmium content in particular exposure pathways - diet, drinking water, ambient air and soil -were processed for the period 1994-2003. The estimate of the daily dietary intake for an average adult population amounted to 11-19 microg/d, i.e. 0.17-0.30 microg/kg bw/d, which represents 17%-30% of the PTWI (provisional tolerable weekly intake). The contribution from drinking water to the oral exposure is low; on average 0.5 microg/d. Potential exposure to airborne Cd was estimated at about 0.02 microg/d. The additional Cd intake from urban soil ingestion probable in small children was found to be insignificant based on Cd concentrations in the soil of kindergarten playgrounds. Biomonitoring outputs characterize the recent and life-long cadmium burden of the Czech population from general environment In 1994-2003, the median blood Cd levels ranged in the interval 0.9-0.4 microg/l blood, in smokers being more than double that in non-smokers. Blood Cd levels detected indicate slightly decreasing trend as well as urine Cd levels (range of median values 0.44-0.28 microg/g creatinine). Since 1996 the levels in children have been found in more than 50% cases below the detection limit of the methods used. The estimated total cadmium intake in the Czech urban population does not signalize any increased risk of health impairment considering non-carcinogenic effects.     

Risks to children from exposure to lead in air during remedial or removal activities at Superfund sites: a case study of the RSR lead smelter Superfund site

Superfund sites that are contaminated with lead and undergoing remedial action generate lead-enriched dust that can be released into the air. Activities that can emit lead-enriched dust include demolition of lead smelter buildings, stacks, and baghouses; on-site traffic of heavy construction vehicles; and excavation of soil. Typically, air monitoring stations are placed around the perimeter of a site of an ongoing remediation to monitor air lead concentrations that might result from site emissions. The National Ambient Air Quality (NAAQ) standard, established in 1978 to be a quarterly average of 1.5 microg/m(3), is often used as a trigger level for corrective action to reduce emissions. This study explored modeling approaches for assessing potential risks to children from air lead emissions from the RSR Superfund site in West Dallas, TX, during demolition and removal of a smelter facility. The EPA Integrated Exposure Uptake Biokinetic (IEUBK) model and the International Commission of Radiologic Protection (ICRP) lead model were used to simulate blood lead concentrations in children, based on monitored air lead concentrations. Although air lead concentrations at monitoring stations located in the downwind community intermittently exceeded the NAAQ standard, both models indicated that exposures to children in the community areas did not pose a significant long-term or acute risk. Long-term risk was defined as greater than 5% probability of a child having a long-term blood lead concentration that exceeded 10 microg/dl, which is the CDC and the EPA blood lead concern level. Short-term or acute risk was defined as greater than 5% probability of a child having a blood lead concentration on any given day that exceeded 20 microg/dl, which is the CDC trigger level for medical evaluation (this is not intended to imply that 20 microg/dl is a threshold for health effects in children exposed acutely to airborne lead). The estimated potential long-term and short-term exposures at the downwind West Dallas community did not result in more than 5% of children exceeding the target blood lead levels. The models were also used to estimate air lead levels for short-term and long-term exposures that would not exceed specified levels of risk (risk-based concentrations, RBCs). RBCs were derived for various daily exposure durations (3 or 8 h/day) and frequencies (1-7 days/week). RBCs based on the ICRP model ranged from 0.3 (7 days/week, 8 h/day) to 4.4 microg/m(3) (1 day/week, 3 h/day) for long-term exposures and were lower than those based on the IEUBK model. For short-term exposures, the RBCs ranged from 3.5 to 29.0 microg/m(3). Recontamination of remediated residential yards from deposition of air lead emitted during remedial activities at the RSR Superfund site was also examined. The predicted increase in soil concentration due to lead deposition at the monitoring station, which represented the community at large, was 3.0 mg/kg. This potential increase in soil lead concentration was insignificant, less than 1% increase, when compared to the clean-up level of 500 mg/kg developed for residential yards at the site.     

Public perceptions of a radioactively contaminated site: concerns, remediation preferences, and desired involvement.

A public attitudes survey was conducted in neighborhoods adjacent to a radioactively contaminated site whose remediation is now under the auspices of the U.S. Department of Energy's Formerly Utilized Sites Remedial Action Program (FUSRAP). The survey's purpose was to ascertain levels of actual and desired public involvement in the remediation process; to identify health, environmental, economic, and future land-use concerns associated with the site; and to solicit remediation strategy preferences. Surface water and groundwater contamination, desire for public involvement, and potential health risks were found to be the most highly ranked site concerns. Preferred remediation strategies included treatment of contaminated soil and excavation with off-site disposal. Among on-site remediation strategies, only institutional controls that leave the site undisturbed and do not require additional excavation of materials were viewed favorably. Cost of remediation appeared to influence remediation strategy preference; however, no strategy was viewed as a panacea. Respondents were also concerned with protecting future generations, better assessment of risks to health and the environment, and avoiding generation of additional contaminated materials.     

Tritium recapture behavior at a nuclear power reactor due to airborne releases

This paper describes the initiatives taken by Cook Nuclear Plant to study the on-site behavior of recaptured tritium released in its airborne effluents. Recapture is the process where a released radioactive effluent, in this case tritium, is brought back on-site through some mechanism. Precipitation, shifts in wind direction, or anthropogenic structures that restrict or alter effluent movement can all lead to recapture. The investigation was started after tritium was detected in the north storm drain outfall. Recent inadvertent tritium releases by several other nuclear power plants, many of which entered the groundwater, have led to increased surveillance and scrutiny by regulatory authorities and the general public. To determine the source of tritium in the outfall, an on-site surface water, well water, rainwater and air-conditioning condensate monitoring program was begun. Washout coefficients were also determined to compare with results reported by other nuclear power plants. Program monitoring revealed detectable tritium concentrations in several precipitation sample locations downwind of the two monitored containment building release vents. Tritium was found in higher concentrations in air-conditioning condensate, with a mean value of 528 Bq L(-1) (14,300 pCi L(-1)). The condensate, and to a lesser extent rainwater, were contributing to the tritium found in the north storm drain outfall. Maximum concentration values for each sample type were used to estimate the most conservative dose. A maximum dose of 1.1 x 10(-10) mSv (1.1 x 10(-8) mrem) total body was calculated to determine the health impact of the tritium detected.     

Urbanisation effects on groundwater chemical quality: findings focusing on the nitrate problem from 2 African cities reliant on on-site sanitation

By 2010 Africa's urban population will have grown to over 420 million with on-site sanitation the predominant excreta disposal option. The use of on-site sanitation has important public health benefits but can result in large faecally derived loadings of nitrogen and chloride to groundwater resources. Nitrate is of particular concern, with elevated concentrations linked to potentially serious health problems. N and Cl can derive from natural sources so it is important to quantify the additional impact of human activities. Several authors have used empirical relationships between nitrate and chloride concentrations to assess the extent to which excreta influences groundwater quality. However, these relationships have assumed fixed loadings from excreta. Relationships between N and Cl have been extended here by adding country-specific estimates of average annual per capita nitrogen and chloride content of, and loading from, excreta. The results are compared with groundwater monitoring results from two very different mid-sized African cities (Timbuktu, Mali and Lichinga, Mozambique) where the vast majority of residents use on-site sanitation and are dependent on the subsurface water for drinking purposes. The results illustrate the impact of urbanisation on groundwater quality. They are compared with data from other African cities to allow the calculation of a general nitrate and chloride relationship for unsewered African urban areas. Potential interventions to help arrest rising nitrate levels and so provide a public health benefit are also examined.     

Dorchester lead-safe yard project: A pilot program to demonstrate low-cost, on-site techniques to reduce exposure to lead-contaminated soil

Despite a general reduction in blood lead levels in children after lead was banned in gasoline and paint, lead poisoning remains an important health problem in many older urban areas. One factor that increases risk in these places is the high levels of lead in certain residential areas. A major intervention study found that reducing lead levels in urban soils results in a reduction in exposed children’s blood lead levels. Removing lead from inner-city soils or reducing exposures to lead-contaminated soils typically is expensive, technologically challenging, or beyond the ability of low-income households to undertake. This project, in conjunction with residents and community-based institutions, developed a series of in situ, low-cost, low-technology measures that worked to reduce the exposure to lead-contaminated soils in one Boston, Massachusetts, neighborhood. The project demonstrated several important results. Government, universities, residents, and community based organizations can work together effectively to reduce exposures to lead in soil. Lead-contaminated soil can be mitigated at a fraction of the cost of conventional methods in ways that increase the ability of residents, community health centers, and others to have a positive impact on their neighborhoods. A lead-safe yard program can be replicated and institutionalized by municipal home de-leading programs and other community organizations.     

Human health risk assessment of lead,cadmium, and mercury co-exposure from agricultural soils in the Tuzla Canton (Bosnia and Herzegovina)

The aim of this study was to assess the risk of human exposure to lead (Pb), cadmium (Cd), and mercury (Hg) through agricultural soil by considering both uncertainty and variability in key exposure parameters. For this reason we collected soil samples from 29 locations in the Tuzla Canton (Bosnia and Herzegovina) and measured their metal levels with inductively coupled plasma atomic emission or absorption spectrometry (ICP-AES and ICP-AAS, respectively). The levels of Pb ranged from 13.33 to 1692.33 mg/kg, of Cd from 0.05 to 3.67 mg/kg, and of Hg from 0.02 to 2.73 mg/kg. To estimate cancer and non-cancer risks we used deterministic and semi-probabilistic methods. Lead was found to involve higher health risk than the other two heavy metals. Its hazard index (HI) decreased between population groups (children>women>men) and exposure routes (ingestion>skin contact>inhalation). Our Monte Carlo simulations indicated that Pb HIs for both adult populations had a 0.6 % probability to exceed the threshold value of 1, while in children this probability was 14.2 %. Cd and Hg showed no probability to exceed the threshold in any scenario. Our simulation results raise concern about possible adverse health effects of heavy metals from soil, especially in children. It is very important to continue monitoring environmental pollution and assess human health risk, not only with respect to soil, but also with other important environmental compartments, such as air and water.KEY WORDS: cancer risk, deterministic methodology, heavy metals, Monte Carlo simulations, non-cancer risk, probabilistic methodology     

Arsenic geochemistry and human health in South East Asia

Arsenic occurs naturally in many environmental components and enters the human body through several exposure pathways. Natural enrichment of arsenic may result in considerable contamination of soil, water, and air. Arsenic in groundwater can exceed values hundreds of time higher than the concentration recommended for drinking water. Such exposure levels indicate a serious potential health risk to individuals consuming raw groundwater. Human activities that have an impact on the environment may increase the distribution of inorganic arsenic. Abandoned mines are of great concern due to the extremely high arsenic concentrations detected in mine drainage and tailings. Diet, drinking water, air, soil, and occupational exposures are all sources of inorganic arsenic for humans. Interdisciplinary efforts to better characterize the transport of arsenic and reactants that facilitate their release to the environment are important for human health studies. Multi-disciplinary efforts are needed to study diet, infectious disease, genetics, and cultural practices unique to each region to better understand human health risk and to design public health interventions.     

The risk to the United Kingdom population of zinc cadmium sulfide dispersion by the Ministry of Defence during the "cold war"

Objectives: To estimate exposures to cadmium (Cd) received by the United Kingdom population as a result of the dispersion of zinc Cd sulfide (ZnCdS) by the Ministry of Defence between 1953 and 1964, as a simulator of biological warfare agents. Methods: A retrospective risk assessment study was carried out on the United Kingdom population during the period 1953–64. This determined land and air dispersion of ZnCdS over most of the United Kingdom, inhalation exposure of the United Kingdom population, soil contamination, and risks to personnel operating equipment that dispersed ZnCdS. Results: About 4600 kg ZnCdS were dispersed from aircraft and ships, at times when the prevailing winds would allow large areas of the country to be covered. Cadmium released from 44 long range trials for which data are available, and extrapolated to a total of 76 trials to allow for trials with incomplete information, is about 1.2% of the estimated total release of Cd into the atmosphere over the same period. "Worst case" estimates are 10 µg Cd inhaled over 8 years, equivalent to Cd inhaled in an urban environment in 12–100 days, or from smoking 100 cigarettes. A further 250 kg ZnCdS was dispersed from the land based sites, but significant soil contamination occurred only in limited areas, which were and have remained uninhabited. Of the four personnel involved in the dispersion procedures (who were probably exposed to much higher concentrations of Cd than people on the ground), none are suspected of having related illnesses. Conclusion: Exposure to Cd from dissemination of ZnCdS during the "cold war" should not have resulted in adverse health effects in the United Kingdom population.     

Selenium (Se) Does Not Reduce Cadmium (Cd) Uptake and Translocation in Rice (Oryza sativa L.) in Naturally Occurred Se-Rich Paddy Fields with a High Geological Background of Cd

This study examined the selenium (Se) and cadmium (Cd) uptake by rice from soil and analyzed the relationship between Se and Cd in naturally occurred Se-rich paddy fields with a high geological background of Cd. Significant correlations were observed between soil Se and plant biomass Se, but not between soil Cd and plant biomass Cd. High concentrations of Cd were detected in rice plants and particularly in rice grains, suggesting potential health risks to human. Contrary to results from other previous studies, our results showed that high soil Se did not reduce Cd uptake by rice, although it decreased the availability of Cd in soil. Rather, soil Se and internal Se pool in rice were positively correlated to the transfer of Cd from root to straw. The effect of Se on the uptake and translocation of Cd in rice in field is therefore different from those in pot experiments.

     

Communities near toluene diisocyanate sources: an investigation of exposure and health

Toluene diisocyanate (TDI) is a well-known cause of occupational asthma, but we know little about the potential for exposure and health effects among residents who live near facilities that release TDI. In the mid-1990's, the North Carolina Department of Health and Human Services and the Agency for Toxic Substances and Disease Registry investigated exposures to TDI and health outcomes in one community, which left some unanswered questions. This cross-sectional study evaluated the potential associations between living near a TDI source and the prevalence of three variables: asthma or asthma-like respiratory symptoms, antibodies specific to TDI, and verifiable levels of TDI in residential air. Results among North Carolina residents living near such facilities (five target communities) were compared with the results from residents living further away (five comparison communities). Overall, the prevalence of reporting either asthma or asthma-like respiratory symptoms was higher (odds ratio = 1.60; 95% confidence interval = 0.97-2.54) among residents in target communities than those in comparison communities. However, this difference was not statistically significant. Symptom prevalence varied greatly among the community populations. The prevalence of respiratory symptoms was higher near facilities with historically higher TDI emissions. Among the 351 participants who provided blood samples, only one had immunoglobulin G specific antibodies to TDI. This participant lived in a target area and may have had non-occupational exposure. TDI was detected at an extremely low level (1 ppt) in one of the 45 air samples from target communities. One ppt is one-tenth the EPA reference concentration. Overall, air sample and antibody test results are not consistent with recent or ongoing exposure to TDI.     

Traffic-related exposures,constrained restoration,and health in the residential context

Traffic-related exposures may undermine the restorative character of the home, and this may in turn undermine health and residential satisfaction. We addressed this possibility with data for adults residing in a large valley near Innsbruck, Austria (N=572). We joined objective measures of traffic-related sound and air pollutants with reports from door-to-door surveys concerning perceived disturbance from traffic-related exposures, restorative qualities of the living environment, self-perceived health and residential satisfaction. We analyzed these data with successive tests of nested structural equation models, with and without the restorative quality variables. The results suggest that the negative impact of traffic-related exposures on self-perceived health and satisfaction with the living environment involves the constraint of restorative qualities of the living environment, over and above the share traditionally attributed to such exposures viewed as stressors. We discuss theoretical and practical implications of the distinction between environmental stressors and constraints on restoration.     

Community exposure to perfluorooctanoate: relationships between serum concentrations and exposure sources

OBJECTIVE: The objective of this study was to determine serum (perfluorooctanoate [PFOA]) in residents near a fluoropolymer production facility: the contributions from air, water, and occupational exposures, personal and dietary habits, and relationships to age and gender. METHODS: The authors conducted questionnaire and serum PFOA measurements in a stratified random sample and volunteers residing in locations with the same residential water supply but with higher and lower potential air PFOA exposure. RESULTS: Serum (PFOA) greatly exceeded general population medians. Occupational exposure from production processes using PFOA and residential water had additive effects; no other occupations contributed. Serum (PFOA) depended on the source of residential drinking water, and not potential air exposure. For public water users, the best-fit model included age, tap water drinks per day, servings of home-grown fruit and vegetables, and carbon filter use. CONCLUSIONS: Residential water source was the primary determinant of serum (PFOA).