A 394 VOLUME 111 | NUMBER 7 | June 2003 • Environmental Health Perspectives

Environews Focus

Body of Evidence

Environmental Health Perspectives • VOLUME 111 | NUMBER 7 | June 2003 A 395

More than a decade after the National Research Council pushed for broad test-ing of toxic substances in people and more than 40 years after the birth ofthe ongoing National Health and Nutrition Examination Survey

(NHANES), data on human exposures to potential toxicants continue to come in.The latest contribution came with the January 2003 release of the Second NationalReport on Human Exposure to Environmental Chemicals by the Centers for DiseaseControl and Prevention (CDC). The report presents findings on the presence of asmall selection of potentially toxic substances in a group of volunteer test subjects.It more than quadrupled the number of substances covered in the CDC’s firstreport released two years earlier, adding in new chemicals, congeners, and natural-ly occurring substances. Already it is a useful tool for many organizations, physi-cians, and agencies.

“The ability to have a database like this as a baseline reference is very powerful,”says Paul Gilman, assistant administrator for research and development in the U.S.Environmental Protection Agency (EPA) Office of Research and Development.“This work puts things in a context for us.”

The CDC findings revealed some significant reductions during the past decadefor some of the substances, such as lead and secondhand tobacco smoke. But thefindings also unveiled many puzzles, such as large variations in some of the con-centrations measured by age, sex, and race/ethnicity, and differences of 200-fold ormore between concentrations found in the lowest and highest percentiles reported.These puzzles highlighted one of the main points made by the CDC with the releaseof the report—that much about human exposure remains a mystery, and that oneof the primary uses of the data will be to help shape future research. The CDCreport offered no new information on health effects from the measured exposures,another field in which the CDC acknowledged much research is needed.

The CDC’s decisions on which substances to include in the Second NationalReport were influenced by six considerations: data showing exposure to U.S. resi-dents; known or suspected serious health effects from exposures; availability of goodanalytical methods to evaluate a substance; testing costs; availability of adequateblood or urine samples; and a desire to track selected public health interventionsover time.

The 116 substances covered in the Second National Report are only the tip ofthe iceberg as far as potential chemical exposures go. About 2,900 high produc-tion volume chemicals—regularly produced or imported in volumes of 1 millionor more pounds per year—are among the chemicals under investigation by manyagencies and companies. There are at least 80,000 chemicals, with largely C

hri

sto

ph

er G

. R

euth

er/E

HP

Focus | Body of Evidence

unknown toxicity, that have ever been reg-istered under the Toxic Substances ControlAct as being in commerce in the UnitedStates (an estimated 20,000 more haveever been in commerce elsewhere in theworld). There may be about 6 millionchemicals that exist in the world, says JohnOsterloh, chief medical officer for theCDC Environmental Health Laboratory,although estimates vary and no one reallyknows. Many of these are not produced forcommerce but rather exist as breakdownand incineration products of chemicalsthat are in commerce. The number ofthese substances that may reside for anylength of time in people is a mystery. TheCDC says its data can’t clarify that num-ber, but that such information is irrelevantat the moment, because scientists can’tevaluate the health consequences of suchcomplex mixtures.

But that and other missing informationneeds to be developed as soon as possible, saysMatthew Cahillane, an environmental healthspecialist with the New Hampshire Depart-ment of Health and Human Services. At themoment, “we’re basically firefighters,” he says.“We’re playing catch-up all the time.”

What’s in There?With the advent of U.S. environmentalprograms in the 1970s, the first efforts togauge the impacts of toxic substances onpeople focused on measuring them indi-rectly—in the environment—in part be-cause that was more technically feasible.But those methods, while a key element inunderstanding the overall exposure picture,provide only rough parameters. They can’ttrack what actually gets into people throughall the potential pathways—such as inhala-tion, ingestion, or dermal absorption—andcan’t evaluate what stays in. Plus, theamount of any one chemical detected in abiomonitoring sample presents only a snap-shot taken at one point in time. The expo-sure reflected may be recent or old, chronicor isolated. A single biomonitoring event isinsufficient for understanding the magni-tude, frequency, or duration of exposure, orthe absorption, distribution, metabolism,or excretion of a substance.

With those limitations in mind, someresearchers began to probe what was actu-ally in people. One of the earliest effortsbegan in 1976 with studies of lead. Thefindings of worrisome concentrations in

many people eventually led to campaignsto remove lead from gasoline and paint. Inresponse to public concerns about pollu-tion that built in the 1980s and 1990s, theNational Academies’ National ResearchCouncil recommended in 1991 that thecountry should create a national programto track chemicals in human tissues. Thathelped lead to the adoption and expansionof several efforts. In 1993, the EPA beganits National Human Exposure AssessmentSurvey, which assessed 46 chemicals in460 participants in 3 U.S. regions. Theresults have yet to be fully analyzed, buthave been evaluated for some associationsbetween levels of environmental exposureand biomarkers.

Many other independent studies,including those performed by companiesinvestigating their own chemicals, have pro-vided clues about various substances. Butthese studies often focus on occupationalexposures to specific chemicals under spe-cific conditions, and thus do not applywhen estimating background exposures forthe general U.S. population. Furthermore,numerous barriers, such as funding limita-tions, variations in study methodologies

A 396 VOLUME 111 | NUMBER 7 | June 2003 • Environmental Health Perspectives

Phot

oDis

c

Focus | Body of Evidence

Environmental Health Perspectives • VOLUME 111 | NUMBER 7 | June 2003 A 397

and data systems, and lack of coordinationbetween government agencies, have stymiedthe development of any extensive informa-tion, the U.S. General Accounting Officeconcluded in its May 2000 report ToxicChemicals: Long-Term Coordinated StrategyNeeded to Measure Exposures in Humans.

The first substantial breach in that grid-lock came 21 March 2001, when the CDCreleased its first National Report on HumanExposure to Environmental Chemicals. Thatreport covered 27 substances, includinglead and 12 other metals, 6 pesticidemetabolites, the nicotine metabolite coti-nine, and 7 metabolites of phthalates, aclass of chemicals used in products such asfragrances, industrial solvents, and flexibleplastic products. For that study, researchersevaluated blood and urine samples fromabout 3,800 representative volunteers whoparticipated in the 1999 NHANES.

On 31 January 2003, the CDC unveiledits Second National Report. The CDCexpects the report to be used in a number ofways. One of the primary uses will be todetermine what substances people areexposed to, at what concentrations, andwhether those concentrations are toxic. Forthe few substances that do have establishedtoxicity thresholds, such as lead, it will alsohelp determine how many people may haveunsafe levels.

As the database grows, it will provideincreasingly reliable information on whatthe typical concentration ranges are forvarious population groups, providingphysicians and public health officials withvaluable perspective on what is “normal” inthe United States. It can also provide keyinformation on the success or failure ofintervention programs, such as antismok-ing initiatives, lead control programs, andbans and other efforts aimed at reducingpesticide body burden.

Most Extensive Assessment So FarThe new report offers a greatly expandedbody of information, including both old andnew data on the initial 27 substances andnew data on 89 other substances or theirmetabolites, including dioxins, furans, poly-chlorinated biphenyls (PCBs), polycyclicaromatic hydrocarbons, plant estrogens, andan expanded list of pesticides. With theadditional data available, based on bloodand urine samples drawn in 1999 and 2000from about 1,700–8,000 people per sub-stance, the CDC has been able to providethe first breakdowns by age, sex, and race/ethnicity. “It’s a giant step forward,” saysJim Pirkle, deputy director for science at theEnvironmental Health Laboratory.

To find out more about the potentialadverse health impacts of the substances

found in the CDC study, researchers will bescrutinizing the findings for years to come.Some of their focus will be on the variationsin exposure by demographic strata thatpopped up in the CDC study.

For instance, urine concentration ofbarium, which occurs naturally and is usedin products such as paint, rubber, andceramics, was about 65% higher in childrenaged 6–11 than in people 20 and older, andwas nearly twice as high in non-Hispanicwhites as in Mexican Americans. For thenaturally occurring plant estrogen genis-tein, found primarily in foods such as soy-beans, Mexican Americans have a meanurine concentration about 65% higher thannon-Hispanic blacks. Although associatedwith health benefits including lowering therisk of osteoporosis and some cancers, phy-toestrogens such as genistein also are sus-pected of reducing reproductive capacityand promoting other cancers.

Mercury concentrations in blood wereabout three times higher in females aged16–49 than in children aged 1–5, and weremore than 40% higher in non-Hispanicblacks than in non-Hispanic whites orMexican Americans. About 8% of women of

reproductive age had mercury levels over theEPA’s reference dose for organic mercury.The health effects of mercury depend onwhether it is a metallic, organic, or inorganicform, and on the timing, route, amount, andlength of exposure. Fetuses exposed to organ-ic mercury may experience damage to neuro-logical systems, and adults exposed long-termcan suffer neurological, pulmonary, and gas-trointestinal damage. Kristina Thayer, a sen-ior scientist with the EnvironmentalWorking Group, an environmental organiza-tion, suspects that something as simple ashigher consumption of canned tuna—aknown source of mercury—may play a part.Mercury also comes from sources such ascoal-fired power plants (whose emissions cantravel many miles), medical devices, dentalamalgams, and cosmetics.

Lead concentrations were still high insome people, but the mean concentrationin blood samples of children aged 1–5 wasabout 17% lower than the mean concentra-tion detected in NHANES III in the early1990s. Males in the latest study had 47%higher concentrations than females.

Chemicals in the same class can beabsorbed, metabolized, and excreted by

Focus | Body of Evidence

Is It Enough to Avoid Exposures?A pilot study by the Mount Sinai School of Medicine, in collaboration with twoenvironmental organizations, the Environmental Working Group andCommonweal, revealed that nine volunteers had an average of 91 toxic sub-stances in their bodies. Out of 210 substances tested—which included polychlo-rinated biphenyls, metals, phthalates, pesticides, and volatile substances—167showed up in at least one person, the groups said with the release of theirreport on 30 January 2003. Each test subject had a total of anywhere from 77 to106 substances in his or her blood or urine. On average, participants had 53 car-cinogens, 58 known endocrine disruptors, 53 chemicals that are toxic to theimmune system, and 55 that are linked to birth defects or abnormal develop-ment. The health effects of chemical combinations, or even single substances,remain poorly understood, however.

The study is not peer-reviewed, and, due to the small sample size, the resultsdon’t reflect what might be found in the U.S. population. But Kristina Thayer, asenior scientist with the Environmental Working Group, is concerned by theresults, in part because the nine volunteers—who included environmentalhealth activists, health care professionals, and a journalist familiar with environ-mental health issues—likely lead “less toxic” lives than many people. “Our folksare pretty savvy and know what to avoid, but it wasn’t good enough,” she says.“Even when you try to avoid exposure, you can’t.”

U.S. residents shouldn’t have to be so savvy, says one critic of U.S. toxic sub-stance policies. “This study merely confirms what we’ve known for a long time,”says Samuel Epstein, a professor emeritus of environmental and occupationalmedicine at the University of Illinois School of Public Health in Chicago andchairman of the Cancer Prevention Coalition—that a wide range of toxic sub-stances are getting into people. He has followed similar studies for decades, andsays that government agencies should have taken more aggressive action yearsago to slash exposures to toxic substances. –Bob Weinhold

people in very different ways. Mono-ethylphthalate, created in the body from diethylphthalate, was present in the urine of adultsaged 20 and older at nearly twice the con-centration found in children aged 6–11.Diethyl phthalate is used in products suchas fragrances, soaps, and hand lotions. Incontrast, mono-benzyl phthalate, created inthe body from benzylbutyl phthalate, waspresent in the urine of children aged 6–11at a concentration more than three timeshigher than that found in people aged 20and older. Benzylbutyl phthalate is used inproducts such as adhesives, sealants, and carcare products.

Along with the differences in age, sex,and race/ethnicity, there were some largedifferences in concentrations between theupper and lower percentiles. In the case of2,5-dichlorophenol, a urine metabolite ofp-dichlorobenzene, which is used in prod-ucts such as insect repellents and deodoriz-ers, concentrations in the 95th percentilewere about 230 times higher than thosefound in the 25th percentile. And withseveral of the plant estrogens, the differ-ences between the 95th and 10th per-centiles ranged from 100-fold to nearly200-fold.

Does Exposure Cause HealthProblems?Large variations between percentiles can becaused for three reasons, says GeorgeLucier, an advisor to the National Toxi-cology Program. First, people are exposedto varying amounts of different chemicals.Second, some chemicals are rapidlydegraded by the body, and the timebetween exposure and sampling is criticalto attaining comparable measures. Third,different people degrade chemicals at dif-ferent rates for reasons related to age, sex,genetics, and diet.

The large differences in the CDCreport raised a red flag for Thayer, whonoted that the few toxic substances that dohave health standards typically have only a10-fold safety margin built into them as ahedge against variations in human vulner-ability and uncertainties in regulatoryassumptions. Current hedge factors maynot address the ranges of exposure foundin the CDC report, she observes, andshould be addressed in future investiga-tions and regulations.

Very little is known about the healtheffects that might be caused by the concen-trations found in the test subjects in theCDC study, or from the combined effects ofthese substances. Some industry representa-tives suggest there may not be much toworry about. “It hasn’t been determinedthat reducing exposure would result in

improved health,” says Jennifer Bian-caniello, a spokeswoman for the AmericanChemistry Council.

Among the studies of potentially toxicsubstances that have been conducted, someare showing adverse health effects at whatare considered low doses. For instance,monoethyl phthalate, at concentrationspresent in our environment, causes DNAdamage in sperm, reported Susan Duty ofthe Harvard School of Public Health andcolleagues in December 2002 in EHPonline. The EPA has established a peer-review panel to assess the so-called low-dosehypothesis, which holds that endocrine-dis-rupting chemicals cause adverse healtheffects at extremely low doses. In a 26March 2002 statement on the findings ofthe panel, the agency said the panel foundthat “there were credible studies supportinga low-dose effect, but that the effects were‘. . . dependent on the compounds studiedand the endpoint measured.’” The panelalso identified credible studies that did notsupport a low-dose effect.

Given all the complex interactions thatstudies suggest can occur with exposure totoxic substances and other stressors, andgiven the role of other poorly understoodfactors such as the avenues and timing ofhuman exposure, information on sub-stances in people alone isn’t a magic bullet.“[The Second National Report] is just asnapshot in the middle of the biologicalstory,” Gilman cautions.

Using the ResultsNonetheless, human exposure data are animportant part of the bigger picture. Forinstance, they likely will help shape thework of one of the developing researchtools—the Environmental GenomeProject—when actual testing of yet-unde-termined substances is ready to begin in thenext five years or so, says RaymondTennant, director of the National Centerfor Toxicogenomics. Any methods adoptedmay be useful for examining human bodyburdens in relation to people who are atincreased risk due to genetic predisposition.

Agencies and organizations trackingchronic diseases also are following therelease of human exposure data with greatinterest, and the CDC report will improvescientists’ ability to design more sensitiveepidemiologic studies, Lucier says. Onelikely topic will be chronic diseases such ascancer, asthma, and diabetes. Such diseasesplay a part in 7 of every 10 U.S. deathseach year, cause major limitations in activi-ty for 25 million Americans, and affect atotal of about 90 million Americans tosome degree, according to the CDCNational Center for Chronic Disease and

Health Promotion. Many chronic diseaseshave suspected environmental links. LauraSegal, director of communications for theTrust for America’s Health, a Washington,D.C.–based advocacy organization, saysthe report will “give us a fighting chanceagainst chronic diseases” by providing someof the first data regarding potential linkageswith toxic substances.

The data are already shaping someresearch efforts. Cahillane says that NewHampshire was going to investigate radonas part of its work with the CDC’s chronicdisease studies, but has decided afterreviewing the Second National Report thatmercury is a better research topic.

To help expand the limited nationaldata on chronic diseases, the CDC receivedits first funding in 2002 for a national envi-ronmental public health tracking network.But it will be many years before the accu-mulating data on diseases and toxic sub-stances in people and the environment willbe meshed. “Each of these huge mountainsof data have to be made much more order-ly and understandable,” says Richard J.Jackson, director of the CDC NationalCenter for Environmental Health.

Despite much missing information,public health officials have already begunto use the CDC data to assess environmen-tal health issues. In their study of theunusually high number of acute lympho-cytic leukemia cases in Fallon, Nevada(where suspected exposures include agricul-tural, military, mining, and naturallyoccurring pollutants), the CDC and theAgency for Toxic Substances and DiseaseRegistry (ATSDR) used the SecondNational Report to put their local data incontext. They found that tungsten concen-trations in Fallon residents were far higherthan the national mean cited in the report,and also found slight elevatations in con-centrations of antimony, barium, cesium,cobalt, molybdenum, uranium, and 7 pes-ticides, as well as detectable concentrationsof 18 PCBs. However, the agencies are say-ing so far that there is no proof that suchexposures led to the leukemia increase,although some research continues.

Community activists in Anniston,Alabama, home to extensive PCB pollu-tion, are comparing the CDC’s data onsubstances such as PCBs to their own data,and say they are confirming their suspicionsthat their residents are being exposed tohigh levels of pollutants. “We have resi-dents with godawful numbers that are a lothigher than what [the CDC] found,” saysShirley Baker, a health consultant toAnniston’s Community Against Pollutionorganization. She adds that the communitygroup likely will use the CDC data in its

A 398 VOLUME 111 | NUMBER 7 | June 2003 • Environmental Health Perspectives

Focus | Body of Evidence

Environmental Health Perspectives • VOLUME 111 | NUMBER 7 | June 2003 A 399

requests for grants and other funding, andin litigation. The ATSDR is developing aproposal for a multiyear, multimillion-dollar study of Anniston, and will use theSecond National Report to help put its find-ings in context, says John Abraham, chief ofthe ATSDR’s Exposure Investigations andConsultations Branch in the HealthAssessment and Consultation Division.

Industry groups, too, are already usingthe CDC studies. “They do inform ourindustry on better ways to do things,” saysAngelina Duggan, director of science poli-cy for CropLife America, a pesticide indus-try organization.

Now What?The CDC plans to keep building its data-base and report the results every two years.The Second National Report cost $6.5 mil-lion in direct laboratory expenses, withother expenses included in the $25 millionthat NHANES spends each year for itsoverall study. The next exposure studyshould also cost about $6.5 million, whichalready has been committed, Osterloh says.

Blood and urine samples for the nextreport, covering 2001 and 2002, have

already been drawn from volunteers, fol-lowing protocols similar to the two earlierstudies. Samples from sources in the bodyother than blood and urine—such as fat orbone—likely would reveal additional bodyburden information, but have not beentaken because they require more invasivetechniques.

The number of substances that will beanalyzed will expand, and may includeabout 30 volatile organic compounds(including some chlorination by-products),arsenic, bisphenol A, polybrominateddiphenyl ethers, perchloroethylene, and anumber of other substances.

Lynn Goldman, a professor of occupa-tional and environmental health in theJohns Hopkins Bloomberg School ofPublic Health, appreciates the data that areavailable, but is looking forward to evenbetter information. “Some of the thingsI’m most interested in are yet to come,” shesays. For instance, body burden data forchildren aged 5 and under would be veryhelpful in her work as a pediatrician.However, that information will be difficultto obtain with the present testing methods,she and the CDC acknowledge, because

blood and urine samples from small chil-dren often aren’t large enough to do exten-sive testing.

Researchers at the CDC are tweakingthe analytical techniques within the threegeneral methods they already use: isotopedilution mass spectrometry, inductivelycoupled plasma mass spectrometry, andgraphite furnace atomic absorption spec-trometry. One benefit to this tweaking willbe lower limits of detection for many sub-stances. For the next report, “some of thedetection limits would improve dramati-cally,” says Osterloh.

Jackson hopes to be able to transferthese continually refined methods to otherlaboratories around the country, becausesuch capabilities are rare right now. Hewould also like to see the data broken outgeographically in the future. “At somepoint, we hope to provide state-by-statedata,” he says. He cautions that in order toacquire enough data, such an effort wouldtake a substantial investment by individualstates. Some states, however, have alreadyexpressed interest in the idea.

Bob Weinhold

Phot

odis

cFocus | Body of Evidence