Human Exposure



Exposure Modeling

Direct sampling measurements (at the time of exposure) of the air we breathe, the water we drink, or the foods we eat are rarely done. Instead, human exposure to toxic substances is often estimated from other data by mathematical exposure modeling.  Data commonly used for exposure modeling include release data (e.g., TRI), ambient environmental monitoring data (e.g. air quality), estimates of contaminated food intake obtained by population-based dietary surveys such as the National Health and Nutrition Examination Survey (NHANES),[1] activity surveys such as the National Human Activity Pattern Survey (NHAPS),[2] biomonitoring measurements taken from groups of people with known exposures, and epidemiologic studies. 

A wide variety of factors (e.g., atmospheric conditions, distance from sources, time spent in various activities, etc.) are entered into exposure models to attempt to simulate conditions that accurately reflect what exposures actually occur for a given individual. Biology, chemistry, physics, and math may all play a role in developing models, and model equations are often quite complex with multiple variables. The validity of exposure estimates produced by model calculations thus depends on both (1), the quality of the data that enter the model, and (2), the accuracy and completeness of the model’s assumptions. A few examples of the uses of modeling to estimate exposure are CHAPIS (the Community Health Air Pollution Information System),[3] Scorecard,[4] and the Food Safety Risk Analysis Clearinghouse.[5]

Community Biomonitoring

Biomonitoring is the direct measurement of people's exposure to toxic substances in the environment by measuring the substances or their metabolites in human specimens, such as blood or urine. Biomonitoring measurements are the most health-relevant assessments of exposure because they indicate the amount of chemicals that actually get into people (from all environmental sources (e.g., air, soil, water, dust, food) combined.  Biological samples that have been successfully used as biomarkers of exposure to environmental pollutants include blood, hair, fingernails and toenails, breast milk, bone, teeth, urine, and feces.[6]

Standard approaches to biomonitoring include blood and urine screening for toxic substances. Common substances tested include PCBs, dioxins, furans (byproducts of PVC production, industrial bleaching and incineration), heavy metals, organochlorine insecticides, organophosphate insecticide metabolites, phthalates (plasticizers), and volatile and semi-volatile organic chemicals such as ethyl benzene. Biomonitoring of workers potentially exposed to harmful chemicals in the workplace is required by law in many industries. Lead screening is the most common form of community biomonitoring for environmental exposures, but this section looks at other examples as well.

The CDC Second National Report on Human Exposure to Environmental Chemicals presented biomonitoring exposure data for 116 environmental chemicals in the civilian U.S. population over a 2-year period from 1999 to 2000.[7]  Scientists from CDC's Environmental Health Laboratory measured chemicals or their metabolites (breakdown products) in blood and urine samples from selected participants in the National Health and Nutrition Examination Survey (NHANES).  The sampling plan followed a complex, stratified, multistage, probability-cluster design to select a representative sample of the civilian, non-institutionalized population of the United States. The sample design included targeted sampling of African Americans, Mexican Americans, adolescents (aged 12-19 years), older Americans (aged 60 years and older), and pregnant women to produce more reliable estimates for these groups. In 2000, targeted sampling of low-income whites was also included. The NHANES protocol includes a home interview followed by a standardized physical examination in a mobile examination center.  The age range for which a chemical was measured varied.  For lead, ages 1 and older yielded a sample size of 7970, whereas for p,p'-DDE (an organochlorine pesticide), ages 12 and older yielded a sample size of 1964.  The CDC reports that this report gives first-time information about exposure levels for many chemicals in the US population.  Future reports can be used to answer questions such as: Are exposure levels increasing or decreasing over time?  Are public health efforts to reduce exposure working?  Do certain groups of people have higher levels of exposure than others? 

The Body Burden Report by the Environmental Working Group[8] tested the blood and urine of 9 adults for even more chemicals than those investigated by the CDC (210) occurring in consumer products and industrial pollution.  Of the 167 chemicals found, 76 cause cancer in humans or animals, 94 are toxic to the brain and nervous system, and 79 cause birth defects or abnormal development.

Local Biomonitoring

The Allegheny County Childhood Lead Poisoning Prevention Program (CLPPP) conducts blood lead screening for children ages 0-6 door-to-door in high-risk communities and at fixed-site locations such as day care facilities, head start programs, and health fairs.  Laboratory testing services are provided by the Allegheny County Division of Laboratories.  Pennsylvania is part of CDC’s Blood Lead Laboratory Reference System (BLLRS), a standardization program designed to improve the overall quality of laboratory measurements of lead in blood. In Pennsylvania, 11 laboratories participate in BLLRS. The program, funded under a grant from the CDC, allows these laboratories to evaluate their performance on laboratory tests, providing materials free of charge four times a year.

Screenings are performed year round and are on-going.  Approximately 4000 screenings occurred last year and there were approximately 300-400 cases identified as having a blood lead level in excess of the guideline of 9mg/dl established by the CDC.[9]  The population of children ages 0-6 is approximately 72,000.[10]  The coordinator identified several difficulties in reaching children in this age group and stated that there is still not enough screening going on in the county.  Difficulties range from the ineffectiveness and inefficiencies of the door-to-door screening method, problems in managing identified cases, and the public’s lack of awareness about the effects of lead on child development.  Improvement in monitoring blood lead levels in the county could be achieved by increasing the number of screenings conducted by private physicians.[11]

Many states, including Pennsylvania, target their screening resources to children considered at highest risk.  This approach makes good use of limited funds but does not necessarily produce data representative of all children aged 1-5 years. Therefore, estimates obtained from state and local surveillance data cannot be directly compared to NHANES.[12] 

The PA State Department of Health maintains the Blood Lead Surveillance System.  Laboratories approved to perform blood analysis for lead are required to report blood lead levels for individuals up to the age of 16 and pregnant women.  In 2002, the DOH reports data for 71,776 children under age 16 screened out of an estimated state population under age 16 of 2.5 million.  Philadelphia County reported the largest number of screenings at 37,637, followed by Delaware, Montgomery, and Allegheny.[13]  Besides lead screening, there is currently little biomonitoring in Pennsylvania. In fiscal year 2001, NCEH (CDC’s National Center for Environmental Health) awarded a planning and capacity assessment grant to Pennsylvania to develop a plan for implementing a biomonitoring program for the state. The goal of the grant was to allow the state to make decisions about which environmental chemicals within its borders were of health concern and to plan for measuring levels of those chemicals in the Pennsylvania population. This grant was not refunded.[14]

Other Examples of Biomonitoring

Other states (e.g., California[15]) and countries have made extensive use of biomonitoring. For example, one approach to measure a community’s exposure to environmental pollutants is to examine breast milk. Sweden’s national breast milk monitoring program led to important advances, such as voluntary and government regulations of flame retardants after the levels in breast milk increased dramatically over a short period.  A Breast Cancer Fund and EPA-funded study in Torrance, the Central Valley, and Marin County, CA began to collect breast milk from mothers of newborns demonstrate, in part, how researchers can conduct biomonitoring in a particular community without alienating its residents.  Legislation was proposed, but was not passed, to develop a community-based participatory program to breast milk biomonitoring (CalBBC).  The Collaborative on Health and the Environment’s Breastmilk Monitoring Discussion Group has addressed how communities can best engage with researchers to make sure their concerns are dealt with and how data might be shared with community members and with members of the media.  One of the concerns with breast milk testing is that women could be deterred from breastfeeding once breast milk monitoring increases public awareness about the kinds and quantities of chemicals found in our bodies.

The future development of biomonitoring as a tool for exposure assessment in community settings will depend on our ability to accurately and cost-effectively test for substances of interest in the body as well as the time, expense, and expertise needed to conduct such programs using sound epidemiological methods. In New York State, for example, a CDC-funded program is allowing the development of several innovative targeted pilot biomonitoring programs. These include (1), a program testing urine samples of children in New York City potentially exposed to mercury in religious rituals, (2), a program testing urine cotinine levels as a marker for exposure to second-hand smoke following a statewide ban on smoking in public places, (3), a program testing angler’s blood serum for PBDEs, (4), a program testing children’ urine for poly-aromatic hydrocarbons as a marker of exposure to diesel exhaust, and (5), a program examining the blood and urine of workers at the World Trade Center after 9/11 for a wide range of toxins, including dibenzofurans, dioxins, PCBs, and trace metals. [16]

Case Study #4: Hair Analysis for Mercury in Environmental Journalists


At the 2004 annual meeting for the Society for Environmental Journalists (SEJ) at Carnegie Mellon University, Dr. Jack Spengler and colleagues at the Harvard School of Public Health conducted a study of mercury exposure using hair as a biomarker.[17]  Dr. David Senn, a co-investigator on the study, explained the principle: “As new hair forms under the scalp, it comes in contact with blood and, consequently, circulating methyl mercury. The forming strands incorporate amounts of methyl mercury that are proportional to the levels circulating in bloodstreams. It takes about 30 days for 1 cm of new hair to develop before it pops through the scalp, and so a hair snipped closest to the head reflects methyl mercury levels in the body from the most recent months.”[18] The study, funded by the Heinz Foundation and in partnership with SEJ and PennFuture, examined hair samples from 260 SEJ conference attendees. Participants also completed surveys, supplying demographic information and quantifying their fish consumption habits.  Results of the study were reported in a special session at the end of the conference and indicated the following:

·         27% of participants had Hg hair concentrations that were greater than the 1 µg/g (1 ppm) level corresponding to the USEPA reference dose (RfD) for methyl mercury, The USEPA defines the RfD as an “estimate (with uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.”

·         The number of fish meals reported consumed was a strong predictor for mercury levels in hair, the lowest levels being generally found in vegetarians and vegans who ate no fish,

·         Age was also found to be a statistically significant predictor of Hg levels, even after accounting for differences in fish consumption habits.

"What awful choices,” said Dr. Spengler, “we are being asked to make–lower your mercury levels by switching away from beneficial fish, instead of reducing the source of mercury with the currently affordable control technologies for power plants. It is not just the U.S. power plants that need to be controlled. When you buy the next item made in China, think of the mercury that manufacturing that product is contributing to the Alaskan salmon. And if you enjoy sport fishing in New England lakes, how do you feel when the advisories warn you not to eat what you catch?”[19] 

This study’s innovative approach appears to be a valuable way of increasing public awareness and involvement in monitoring human exposure to environmental toxins. Because hair sampling is relatively easy and non-invasive, participation rates for this type of study are generally high. Moreover, the study brought to light the lack of pre-existing data and public awareness about mercury exposures in the local population, provoking questions from participants such as: To what other toxic substances am I being exposed?  How can I protect myself from being exposed to these substances? and Is my government doing enough to reduce the presence of these chemicals in my environment? Beyond the important information it can yield, the direct measurement of toxins in people can be a powerful impetus for increasing protective individual behaviors as well as more generating effective public health policy.