sign in sign up
<<
Home , Biomonitoring

Human Biomonitoring of Environmental Chemicals

Measuring chemicals in human tissues is the "gold standard" for assessing people's exposure to

Ken Sexton, Larry L. Needham and James L. Pirkle

hat chemicals in your d aily rou­ mine the precise agent, the details of sure also requires complex detective W tine should you be most con­ contact and the full extent of the affect­ work to discover all kinds of d etails, ce rn ed about? Th e vo latile orga ni c ed population. Complicating matters, incl uding the chemical identity (for ex­ compounds from your ca rpet? The ex­ th e scie ntific unde rstanding of the ample, the pesticide chl o rpy rifos), haust fumes on the road to work? The mechanisms of exposure, sud1 as how source (nearby agricultural use), medi­ pesti cide residues in the apple in your various compoWlds are ca rried through um o f transport (ground water) and lund ,? Most of us are exposed to low the air and changed along the way, is route (drinking contaminated well wa­ levels of thousands of toxic chemicals oft en incomplete. As a result, epidemi­ ter). Scientists must consid er this infor­ every day. How can a person-or a na­ ologists oft en find it difficult to estab­ ma tion on exposure against the back­ tion-decide which substances should lish cause-and-effect relationships for ground of people's acti vity patterns, be controlled most rigorously? environmentally induced sickn esses. eating and drinking habits, and lifestyle, One strategy is to go after the largest With out reliable information some pol­ and they must also evaluate the influ­ sources of pollution. This approach cer­ lutants may be wlfairly blamed, where­ ence of other d1emicals in the air, water, tainly makes sense when those pollu­ as others exert their dire effects w ithout beverages, food , dust and soil . OveraU, tants have obvious and w id espread cha llenge. Fortunately, there is hope: a this is a daunting challenge. consequences, such as warming th e method of accurately measuring not Historica ll y, those scientists w ho un­ globe, causing algal blooms, eroding the only contact with, but also absorption dertook such a complex task have re­ ozone layer or killing off wildlife. But of toxic chemicals from, the environ­ li ed on indirect me th ods: question­ for protecting human hea lth, this strate­ ment- human biomonitoring. naires, diaries, interviews, centralized gy does not serve so well, because the monitoring of comn1uluty air or water, link between a given compolUld and its Is It in Me? and a record of broad activity patterns bio logical effects can be difficult to Each person's risk of developing an en­ among the population. But th e results gauge. For epidemiologists to correlate vironmentall y related disease, such as were o ft en disappo inting. Although envirOllmental po llutants w ith health cancer, results from a unique combina­ these ci rcumstantial approaches have problems, they need to know who has tion of exposure.. genes, age, sex, nutri­ the ad vantages of practicality and fru­ been exposed and at what level. ti on and lifestyle. Science doesn't fully gali ty, th ey can also introduce substan­ This knowledge is exceptionally dif­ understand how these variabl es inter­ tial uncertainty into resulting exposure fi cult to gain when there is a lag be­ act, but exposure is cl ea rl y a key fac­ estimates. Tltis shortcoming muJtiplies tween exposure and the manifestation tor. Thus, a fundamental goal of envi­ the potential for a fundamental error­ of illness. In such cases, th e data are ronmental health policy is to prevent classifying a person as "not exposed" seldom- if ever- sufficient to deter- (or at least reduce) people taking in when he or she has been or vice versa. chemicals that to any of the fi ve A second approach, the direct mea­ Kf?II Sextoll is n professor of ell uirOllmelltni sci­ Ds-discomfort, dysfuncti on, disabili­ surement of an individual's environ­ ellces at the Ulliversity Of Texas Sc/wo f of Public ty, disease or d eath. ment, is sometimes a possibility-for ex­ Health, Browl/suifle Regiollal Campl/s, alld past Exposure to an e nv ironmental ample, a person might ca rry a portable presidellt of the illtematiollal Soc iety of Exposure chemi ca l is minimally defined as con­ monitor to record contact with airborne Allalysis (lSEA) . iJ1rry L. Needham is Chief of the tact w ith the skin , mouth or nostrils-a chemi cals. Al though this technique of­ Orgallic Allalytical Bral/ch ill the Na ­ meanjng tha t includes breathing, eat­ fers an unequivocal record of cheln ical tiolla/ Cellter fo r Ellvirolllll(?lltal Health of the ing and drinking. For the purposes of contact, it is technologica lly infeasible Cellters for Disease COlltrol alld Prevelltioll assessing risk, the most im portant at­ or prohibitively expensive to measure (CDC) alld the CIIrrellt lSEA pn'Sidellt . James L. tributes of exposure are magnitude most pollutants this way. Also, although Pirkle is the Deputy Director fo r Sciellee at the CDC's ElluirOllm(?lltal Health LAboratory. Sex­ (what is the concentrati on?), duration sud , monitors document exposure, they tOl" S address is Ull iversity of Texas School of Pub­ (how long does contact last?), frequen­ tell nothing about the person's uptake lic Hmlt/I , Browl/suille Regiollal Campus, RAHC cy (how o ft en d o exposures occur?) o f these airborn e che mi ca ls- how Bllildillg, 80 Fort BroWlI , Browl/suille, TX 78520 . and tim.in g (at what age do exposures much truly gets into his or her body, Illfer/ wt: ksexfOl'@ll fb.edll occur?). The calculation of actual expo­ w hich is, of course, th e most relevant © 2004 Sigma Xi, The Scienti fic Research Society. Reproduction 38 American Scientist, Vol ume 92 with pennission only. Contact [email protected]. . '::: \ -. , , .

•

• ..

Bettmann/ Corbis Figure 1. In July 1945, DDT was widely (and mistakenly) hailed as a progressive measure to eradicate disease-bearing mosquitoes without pos­ ing a ri sk to human health. In this photo from a beach on Long Island, New York, a new insecticide-spraying machine is tested as beachgoers play in the mist. Although this chemical contact is obvious, many other sources of environmental chemical exposure are more difficult to iden­ tify. Human biomonitoring examines people's blood and to evaJuate actual levels of more than a hundred substances. information for assessing health risk. first applied about 130 years ago when dose, or sometimes ilie body burden) Fortwlately, technological ad vances in doctors mon..itored the amount of sali­ has considerable value for estinlating biomedicine and cyluri c acid in the urine of rheunlatics the risk to health. Today, it is relatively now make it possible to get exactly tills who were being treated with large dos­ afford able to m easure the absorbed information. Biolllonitoring meaSUTes es of salicyli c acid (ilie precursor of as­ doses for hu ndreds of chemicals by the actual levels of suspected environ­ pirin). And as early as the 1890s, fa ctory looking for of exposure in mental chemicals in hwnan tissues and workers who were exposed to lead had accessible human tissues and fluids, in­ fluids. This third approach has come to their blood and urine screened to fore­ cluding saliva, semen, urine, sputum, be ilie "gold standard" for assessing ex­ stall ilie elevated levels that produced hair, feces, . and fingernails posure to chemicals. acute . (all of w hich can be coll ected readily), These investi ga tors soon lea rn ed and blood, lung tissue, bone marrow, Blood (and Urine) Will Tell that the degree of contact with a sub­ follicula r fluid, adipose tissue a nd Biomonitoring is not new. It has its roots stance doesn't necessarily determine blood vessels (which require incursion in ilie analysis of biological sanlples for the biologically relevant exposure to into the body). Although procedures to markers for various pharmaceutical that ch emical. As a result, th.i s measure coll ect any of the first set would, tech­ compounds and occupational chemi­ didn't help much in predicting the nica Uy, be considered "noninvasive," cals, efforts that sought to prevent the risks of lead poisoning. H owever, they in fact, that categorization rests on cul­ harmful accumulati on of dangerous did find that the a mount of a com ­ tural, psychological and social factors. substances. A1tllOugh it had a different pound tl,at crosses the body's bound­ So obtaining the ri ght material can name at the time, the general idea was aries {ca ll ed the i.nternal or absorbed sometimes be awkward. Fortunately www.americanscientist.org 2004 January-February 39 natural compounds in the body, such as protei ns. Choosing the appropriate tissue or oXlcan toxicant 2 fluid fo r biological monitoring is based primarily on the chemical and physical properties of the chemical of interest and, in some cases, th e time interval emission since th e last exposure. For example, source some chemica ls including diox in s, polychlorinated biphenyls and organo­ chlorine pesticides have long biological resid ence times in the body (months or years) because they are sequestered in fatty tissues. They are thus said to be fat-l oving or, to use the proper term, • • l lipophilic. By contrast, other chemicals • • • pathway ---­ such as organophosphate pesticides • • . ~ • •• and volatile organic compounds, which • don' t accumulate in fats (being li po­ phobic), have relatively short biological residence times (hours or days) and tend to be metabolized rapidly and ex­ creted in the urine. 1 The time since the last exposure can potential also playa key role in detemlining the dose best biological specimen for analysis. For example, a persistent chemical, such ~ as a dioxin, remains present in blood for a much longer period (years) than does I a nonpersistent compound such as ben­ absorption barrier zene (hours), but dioxin does not form Significant urinary metabolites, whereas does. For these reasons, persis­ .­ tent chemicals are typically measured in blood, and nonpersistent chemi cals are internal measured in urine (as soon after expo­ sure as possible), although they can also dose be detected in blood soon after ex po­ sure if the analytical methods are suffi ­ ciently sensiti ve-and they usually are. Specia lists ca n now detect extremely low levels--parts-per-billion, parts-per­ trillion, even parts-per-quadrillion-

40 America n Scientist, Volume 92 Because the science underpinning food soiVdust water human biomonitoring has improved levels levels levels air significantly in recent years, these ques­ tions are now easier to answer. The nu:;~:~~al jj/levelS rapid advancernent in knowledge of heallh w hat the body does to chemicals that are inhaled, ingested or absorbed through the skin has led to better inter­ hleslyle ~mathematlcal predicted level of toxicant pretation of the range of concentrations personal • modeling ---+ in people for various biomarkers. And the num­ ber of testable compounds has in­ hag::etic ~t ~ creased dramatically: Sensitive and spe­ predisposition /~ cific biomarkers are ava ilable for man y environmental che micals, including lung , intestine and' i... skin absorption rates excretion metals, dioxins, furans, polychlorinated metabolism accumulation biphenyls, pesticides, volatile organic compounds, phthalates, phytoestrogens Figure 3. Traditional estimates of human exposure have to account for many variables, in· and environmental tobacco smoke. As eluding some that demand assumptions about factors that are poorly understood. The resu lt is research continues, the list will surely often uncertain. continue to grow.

Exposure mId Uptake Biomonitoring has many advantages human tissues over traditional methods. For example, or fluids biological s.unples reveal the integrated effects of repeated contact. Also, this approach documents all routes of expo­ personal sure-inhalation, absorption through the skin and ingestion, including hand­ to-mouth tra nsfer by children. Such specimens also reflect the modifying in­ fluences of physiology, bioavailabili ty and bioaccumulation, which can mag­ environment or nify the concentrations of some envi­ microenvironment ronmental chemicals enough to raise them above the detection threshold. Perhaps most importantly, these tests ~ can help establish correla tions between exposure and subsequent illness in in­ dividuals-which is often the key ob­ servation in proving whether or not a emission link exists. A great strength of biomonitoring is th at it provides Wlequivocal evidence ~ that both exposure and uptake have tak­ en place. In some cases these data can confinn the findings of traditional expo­ sure estirnates. For exam ple, in 1979, res­ idents of Triana, Alabama, were noti­ fied that fish from a nearby creek had forty times more DDT than the allow­ able limit, even though the local DDT manufacturing plant had been inactive since 1971. The announcement was es­ pecially concerning because many peo­ ple in that area caught and ate the fi sh regularly. In response to this discovery, the Centers for Disease Control and accuracy Prevention (CDC) constructed an eval­ uation based on DDT concentrations in Figure 4. Exposure to environmental chemica ls can be assessed in several ways. Generally, the fish and the amount of fish eaten per accuracy and cost vary togeth er. Monitoring emission sources is the least expensive and least week. This estimate indeed correlated accurate means of determining human exposure, whereas biomarker measurement is more with levels of DDT and its metabolites, costly but also highly informative for that person.

www.americanscientist.org 2004 January-February 41 gasoline lead decreased by approxi­ mately 55 percent) there was a parallel decline in the amolmt of lead coursing through the veins of the U.S. popula­ tion. Overall, average blood concentra­ tions decreased from about 16 to less than 10 micrograms of lead per deciliter of blood. These data demon­ strated the effectiveness of removing lead from gasoline, and they were a dominant factor in the decision by the Environmental Protection Agency (EPA) to remove lead fro m gasoline more rapidly-a task th at was effec­ tively complete by 1991. Today, the av­ erage blood-lead level in the U.S. pop­ ulation is less than 2 micrograms per deciliter

Exposure Disclosure The study that revealed the tight con­ nection between U,e lead in people's gas tanks and the lead in their blood was mounted by the CDC, which conducts the National Health and Nutrition Ex­ Figure 5. At its Laboratory, CDC scientists use several types of high­ amination Surveys (NHANES for resolution to analyze human tissue and fluid samples. The equipment shown here is being used to measure dioxin levels in a sample of blood serum. (Photograph short). Although no environmental courtesy of James L. Pirkle.) chemicals were measured as part of NHANES I (1971-1975), starting with DOE and DOD, in the blood of Triana to hi gh levels of dioxin. The Air Force NHANES II (1976-1980), the CDC be­ res id ents. In a similar story that un­ first estinlated the risk to soldiers using gan measuring blood lead levels in the folded in U,e late 1980s, chemical-plant a scena rio approach, which included U.s. population, ironically enough, af­ workers in New Jersey and Missouri the average dioxin concentra ti on in ter the Food and Drug Administration discovened that they had been exposed Agent Orange, the number of gallons voiced concerns about possible expo­ to dioxin-contaminated compounds up used during a soldier's tour of duty, sures from eating food stored in lead­ to U,e early 1970s. They had come into and the frequency and duration of p0­ soldered callS, which turned out to be a contact wi th th e dioxin in various tential contact based on job descriptiOll. very minor risk com paned with leaded ways-breathing it, swallowing it and Despite a considerable scientific effort gasoline. As part of NHANES II, the taking it in through the skin. Despite that went into these predictions, CDC EPA tested for certain persistent pesti­ the complexi ti es of their interaction studies in the late 1980s proved that cides in people's blood and nonpersis­ w ith this dangerous substance-and none of the exposure estimates were tent pesticides or their metabolites in the time interval s ince exposure-a correlated with the measured blood urine. After an eight-year hiatus, scheme that used occupational records levels of dioxin in at-risk troops. A sub­ NHANES III was conducted in two to calculate the duration of potential sequent investigation of personnel three-year phases from 1988 to 1994. Tn exposure was able to accurately esti­ with the highest dioxin levels did iden­ th at iteration, the COC measured lead mate internal doses. This finding was tify some patterns that explained their and and began testing for co­ confirmed by the correlation of these increased con tact- for example, small­ tinine, the major metabolite of nicotine, results with the concentration of djox­ statu red enlisted men o ften climbed in blood. Additionally, U,e CDC began a ins in their blood. into the chenlical tanks to clea n out separate pilot program to measure new Having information about exposure residual Agent Orange. compolmds, testing for trace amounts alld uptake is more than a pro forma de­ A more striking example of the val­ of 32 volatile organic chemicals in blood tail: There are many cases in which tra­ ue of biomonitoring came in the mid­ and 12 pesticides or their metabolites in ditional estimates of exposure (q ues­ 1970s when the United States elected urine from approximately 1,000 of the tionnaires, proximity to sources, to start phasing out leaded gasoline. NHANES III participants. environmenta l concen trations, con­ Prior to this decision, traditional mod­ Then came another long gap in cov­ structed scenarios) are not correlated els had suggested that eliminating lead erage. But thankfully, in 1999, NHANES with measured biomarkers. For exam­ in gasoline would have only a slight ef­ became a continuous survey of the ple, fro m 1962 to 1971 , the U.s. Air fect on people's uptake of that metal. non institutionalized U.s. population. Force sprayed the defoliant known as However, biomonitoring data from the (It is thought that excluding members "Agent Orange" in Vietnam. Many ser­ CDC's Second ational Health and of isolated organizations, such as mili­ vice members who participated in that Nutriti on Examination Survey re­ tary personnel, college students and operation touched or breathed the her­ vea led th at from 1976 to 1980 (as un­ prisoners, provides a better cross-sec­ bicide, potentially exposing themselves leaded fu el was first introd uced and tion of America.) In th e cu rrent design,

42 American Scientist, Volume 92 Identifying priority Recognizing time trends in exposures. Out of thousands exposure. Periodic measurement of of chemicals, wh ich are the most biomarkers in the population shows how body dangerous? Biomarkers can help set burdens of chemicals vary from season to priorities for season, year to year and decade to decade. and regulatory C fo llow-up. 1

•

Identifying at-risk Establishing reference populations. Large biomarker studies ranges for comparison. can distinguish exposure differences among A blood test shows that you 've been racial, geographic or socioeconomic groups. exposed to some chemical. Should you be worried? Your doctor can 't tell without data from people with little or no exposure.

Providing integrated Evaluating exposure dose measurements. prevention efforts. Biomarker analysis provides a Our government is entrusted with reducing direct assay of body burden that people's exposure to environmental integrates exposure from all sources, chemicals. Do th ey succeed? Before-and-after even ones that are hard biomarker tests can tell. to measure. 1

Figure 6. When used to establish levels of human chemical exposure, biomonitoring has six major uses that can help to protect public hea lth . www.americanscientist.org 2004 January-February 43 110 17 ::::r 70 c :g, "'Jl 100 Predicted blood lead 16 "t;; 60 '0 c "0 90 15 ::; '"c "0 50 1 - '"~ 14 :9 '"0 80 o;'" j 40 ­ E- 13 > w 70 .!!! g> C "0 :a 12 ~ 30 I 60 8 I '"0> :0 ~ '" 11 c . ~ '" 20 "0 50 '" w "E ] 10 "0 ~ 10 '" 40 .!!! "0 9 '" .!!!'" 30 ~ 0 , 1975 1976 1977 1978 1979 1980 1981 1965 1970 1975 1980 1985 1990 1995 year year

Figure 7. Leaded gasolin e began to be phased out in the 1970s. Although the predicted effect on blood lead was minimal, actual lead exposure in the U.S. population (measured in micrograms of lead perdedlitcr of blood) sharply declined between 1976 and 1980, paralleling the changes in gasoline (left). Blood lead and gas lead continued to fo llow nearly identical decreases up to 1990. At the same time, a series of studies on lead toxicity showed that lower doses could still cause adverse effects, prompting a steady decline in the level defining lead poisoning (right).

a new national sample is collected diox in s, furans and polychlorinated norms, epidemiologists can confirm every two years. Although some other biphenyls; and phytoestrogens. Results the unus ual exposure, identify the studies have locused on specific popu­ from the general population are subdi­ sources and provide continuing health lati o ns or on more restricted data, vided by age, gender and etluticity. ca re as appropriate. The reference NHANES is the only national survey An important feature of the CDC re­ ranges provide indirect financial ad­ that includes both a medical examina­ port is that it provides reference ranges 700 tion and collection of biological sam­ for exposure among the general U.S. ples from participants. Individuals se­ population. If people are concerned ai :g: 600 lected for NHANES are representative that they may have been excessively We ~ u:; 500 of the U.S. population, meaning that exposed to an environnlental chemical, .;:: CI they do not necessa ril y have high or th ey ca n compare their biomarker lev­ Cii CI 400 EE unusual exposures. About 5,000 partic­ els to those standards. These reference g 2 300 ipants are examined annually from 15 ranges are immensely beneficial to 0> :Jl locations throughout the country. public-health scientists who must de­ 200 cid e il certain high-exposure groups 100 .L-'--'-----'--'--=~_ Reporting For Duty need foU ow-up action. Ifaverage levels In March 2001 , the CDC released the among the cohort are similar to those 0.5 National Report on Human Exposure of th e general public, then the group's to En vironmental Chemicals, which in­ exposure is unlikely to cause unique cl uded data from 1999 on 27 chemicals. problems. On the other hand, if levels 0.4 A second report was published in Jan­ are substantiaUy hi gher than national ua ry 2003 that examined 11 6 chemicals 0.3 in samples from 1999-2000. Both stud­ Figure 8. One important function of biomon· ies used biomonitoring to provide an itoring is that it can identify specific subpop­ 0.2 ongoing assessment of exposure to a ul ations that may be more vulnerable to ex· posure from a particular chemical. For variety 01 substances. Although vari­ 0.1 -'---'--'-----'---'--- -'-­ example, p,p'-DDE, a long-lasting metabolite ous studies of workplace exposure, lor of DOT, is more than twice as high in Mexi· example, had rai sed concerns about can-Americans compared with the general 3.0 the health effects of such chemicals, population. By contrast, cotinine levels are most of them had never before been the lowest among this group, indicating that measured in a representative slice of they have th e least exposure to environmen­ the U.s. population. tal tobacco smoke. For both cotinine and lead, The inventory of tested substances in non-hispanic blacks showed the highest lev­ th e second CDC report includes lead, els. DOE (in nanograms per gram of lipid) mercury, cadmium and other metals; and lead (in micrograms per deciliter of persistent (organochl orine-based) and blood serum) data are from th e CDC's Sec­ ond National Report on Human Exposure to nonpersistent (o rganophosphate- and Environmental Chemicals, published in 2003. carbamate-based) insecti cides, herbi­ Information on cotinine (in nanograms per cides and other pesticides; pest repel­ milliliter of blood) is from the third National lents and disinfectants; cotinine; phtha­ Health and N utrition Examination Survey lates; polycyclic aromatic hydrocarbons; (NHANES III), 1988-1991.

44 American Scientist, Volume 92 va ntages too, because distinguishing 5 common from unusual chemical con­ tact helps direct resources to the most­ pertinent exposure situations. 4 nonsmokers smokers The overarching purpose of these re­ ports is to help scientists, physicians and health offi cials to prevent, reduce and treat envirOlIDlentally induced ill­ nesses. However, some caution must be exercised in interpreting the findings: It is important to remember that detect­ in g a chemical in a person's blood or urine does not by itself mean that the ex posure causes disease. Separate sci­ entific studies in animals and hum ans are required to determine wh.ich levels are likely to do harm. For most chenu­ ca ls, tox icologists simply don't have this information. But even if scientists are not sure of 0.1 1.0 10 100 1000 the overall level of risk, they can make serum cotinine (ng/mL) concrete statements about whether sit­ Figure 9. U.S . population clearly segregates into smokers and nonsmokers based on the level of uations are getting better or worse. The cotinine in blood. The working th.reshold for distinguishing the two groups is ]O nanograms la test CDC report, in addition to listing per milliliter of blood serum. Among nonsmokers, the highest values of cotinine were found current biomarker levels in the popula­ in children under 12, and they were strongly reflective of the number of smokers in the home. tion, also highl ights some interesting The data are from NHANES JII, 1988-199]. exposure trends gleaned from earlier NHANES findings. For example, from lations, such as the laborers who apply Bibliography 1991 to 1994,4.4 percent of children be­ pesticides to crops, people li ving near DeCaprio, A. P. 1997. Biomarkers: coming of tween the ages of one and fi ve had lev­ hazardous-waste sites and workers in age for environmental healt h and risk as­ els of blood lead greater than or equaJ lead smelters, all of which are likely to sessment. Ellvirolllllelltal Sciellce & Tee/molo­ gy 31 :1837-1848. to 10 micrograms per deciliter, the Fed­ have higher-than-average exposures to Mendelsohn, M. L., j. P. Peeters and M. J. Nor­ eral action level. By the second collec­ certain environmental chemicals. mandy, eds. 1995. Biomarkers alld Occupa­ tion period (1999 and 2000), only 2.2 tiollal Healtll: Progress and Perspectives. Wash­ percent of this age group exceeded this Annual Check-Up With Biomarkers? ington, IX: joseph Henry Press. threshold . Tim decrease suggests that As th e 21st century unfolds, the CDC Mendelsohn, M. L. , L. C. Mohr and j. P. efforts to reduce lead exposure for chil­ surveys and other well -designed bio­ Peeters, eds. 1998. Biomarkers: Medical and Workplace Applicatioll s. Washington, DC: d ren have been successful. It also serves monitoring studies will continue to joseph Henry Press. as a reminder that some children, in­ build an understanding of people's ex­ Needham, L. L., and K. Sexton. 20C10. Assessing cl ud ing th ose li ving in homes with posure to toxic environmental chemi­ children's exposure to hazardous environ­ lead-based paint or lead-contaminated cals. Nonetheless, these data will not menta l chemicals: An overview of selected dust, remain at unacceptably high risk. obviate the need to coll ect oth er kinds research cha llenges and complexities. /our­ The last report also indicates a hope­ of relevant informati on- to monitor lIal of Exposure Analysis alld Etfviroll1"elltal 10 (Part 2):611-629. ful trend in the exposure to environ­ sources of pollution, to conduct sur­ Needham, L. L. , D. C. Patterson, Jr., V. W. mental tobacco smoke, as shown by veys of toxic substances in the environ­ Burse, D. C. Paschal, W. E. Turner and R. H. tests for th e biomarker cotinine in the ment and to study human acti vities Hill, Jr. 1996. Reference ra nge data for as­ blood of nonsmokers. Median leveJs of and behaviors that contribute to expo­ sessing exposure to selected environmenta l cotinine fell more than 70 percent in sure. Moreover, further research in tox­ toxicants. Toxicology alld Illdu stria l Heaflll 12:507- 513. roughly a decade-that is, between the icology and epidenuology is necessary Pirkle, J. L., E. J. Sampson, L. L. Need ham, D. G. second (1988 to 1991) and third (1 999 befo re s pecialists ca n interpret the Patterson, Jr., and O. L. Ashley. 1995. Using and 2000) periods of data collection. health Significance of ex posure bio­ biological mon..itoring to assess human expo­ This drop provides objecti ve evidence markers for most environmental chem­ sure to priority toxicants. Ellvirollmelltal of reduced exposure to environmental icals. Particularly as detecti on methods Healtll Perspectives 103 (supplement 3):45-48. tobacco smoke for th e general U.s. pop­ improve-enabling investi gators to ulation. Nevertheless, th e fact that more rneasure lower concentrations of more than half of American youth continue chemical s from smaller samples at less to be exposed to environmental tobacco cost-scientific understanding of what For relevant Web lin ks, consult this smoke remains a public-health concern. the body does to the chenucal (and vice issue of America" Scielltist Ol/Iille: The CDC plans to release future re­ versa) must keep pace. If this effort is http·//wwwameticanSCientjs! m:g I ports that document their biomonitor­ successful , a full screen of exposure ISSueTOC/issye/ S21 ing efforts every two yea rs. In the next bio markers may be a part of every edition, they will also add the findings routine physical exam in the not­ from separate studies of special popu­ too-distant future. www.americanscientist.org 2004 jan uary-February 45