Human Biomonitoring of Environmental Chemicals pot

Human Biomonitoring of Environmental Chemicals Measuring chemicals in human tissues is the "gold standard" for assessing people's exposure to pollution Ken Sexton, Larry L. Needham and James L. Pirkle W hat ch emicals in yo ur daily rou- tine should yo u be m os t con- ce rn ed abo ut ? The vo latile or ga nic co mp o und s fr om yo ur ca rp et? The ex- haust fumes on th e ro ad to work? The pes ti ci de residues in th e apple in your lund ,? Most of us are exposed to l ow l eve ls of thousa nd s of toxic chemicals every day. H ow can a person -o r a na- ti on -deci de which s ub stances should be co ntr olled most ri go rousl y? One strategy is to go after the largest so ur ces of pollution. This approach cer- tainly mak es se n se when th ose pollu- ta nt s h ave obvious a nd w id es pr ead co n se quen ces, such as war min g th e globe, causing algal bl ooms, eroding the ozo ne l aye r or killing o ff wi ldlife. But fo r p ro t ec tin g human h ea lth , this s trat e- gy does not se rve so we ll , because the link betw ee n a given compolUld and its biol og ical effects c an be diffic ult to gauge. Fo r epide mi ologists to correlate env ir Oll me ntal pollutants with health pro bl ems, they need to kn ow who has been exposed a nd at what leve l. This kn ow l edge is exceptiona ll y dif- fi c ult to gain when there is a lag be- t wee n ex po sure and th e manif es tation of illness. In such cases, th e data are se ldom- if eve r- sufficient to deter- K f?II Se xt o ll is n p rofessor of e ll u ir O llm e lltni sc i- e ll ces at th e Ulli vers ity Of T exas Sc / wo f of Publ ic He alth , Br ow l/ su ifl e R eg io llal C ampl/ s, alld pa st pr esi de llt of th e illtematiollal Soc iety of Ex pos ur e Allalysis (lSEA) . iJ1r ry L. N ee dham is Chie f of the O rgalli c Allalytic al T ox ic olo gy Bral/ ch ill th e Na - tiolla/ Cellter fo r Ell v irolllll(?lltal He alth of th e Ce llt ers for Di se a se COlltrol alld Pr eve lltioll (C D C) alld the CIIrr e llt lS EA p n'S id e llt . Jam es L. Pi r kl e is th e Deputy Dir ec tor fo r Sci e ll ee at th e CD C's Ell ui rO llm(?llt al He alth LAborat o ry . Sex- tO l" S addr ess is Ull ive rs it y of Te xa s Sch oo l of Pub- li c Hmlt/I , Bro w l/ su ill e Re gio llal Campu s, RA HC Bllildill g, 80 Fort Bro W lI , Br ow l/ su ill e, TX 78520 . Illfe r/ wt: kse xf Ol' @ ll fb .e dll mine th e preci se age nt, th e deta il s of co nt act a nd the full exte nt of the a ff ect- ed po pul a ti o n. Co mplicating matters, th e scie ntif ic und ersta ndin g of the m ec hanisms of exp os ur e, sud1 as h ow v ari ous compoWlds are ca rried through the air a nd chang ed along the way, is o ft en in co mplete. As a result, epidemi- ol og ists o ft en find it diffic ult to es tab- li sh cause- and -e ff ect rela ti ons hip s for env ir onmentally indu ce d sic kn ess es. With out relia bl e info rm a ti on so me pol- lutants may be wlf airly blamed, where- as others exe rt their dire e ff ects without cha ll enge. Fo rtun atel y, there is hope: a method of accurately measuring not only co ntact with, but al so ab so rption of toxic chemicals fr om, th e environ- ment- human biomonito rin g. Is It in Me? Each person's risk of d eve loping an en- v ir onme nt a ll y related disease, such as cancer, results fr om a unique combina- ti on of expos ur e ge nes, age, sex, nutri- ti on a nd lifestyle. Sc ience doesn't fully understa nd ho w th ese varia bl es inter- act, but exp os ur e is cl ea rl y a key fac- tor. Thus, a fundamental go al of envi- ronmental health policy is to pr ev ent (or at least redu ce) p eo ple ta king in chemicals that le ad to any of the fi ve D s-d iscomfort, dysfunc ti on, dis abili- t y, di se a se or de ath . Exp os ure to an env ir o nm e ntal chemi ca l is minimally defined as con- tact with th e skin, mouth or n os tril s-a meanjng that includes breathing, eat- ing and drinking. For the purpo ses of a ssess in g ri sk, th e m os t im po rtant at- tribut es of ex p os ure are m ag nitud e (w hat is th e c on centra ti on?), dura ti on (h ow long does co nt act last?), fr e qu en- cy (h ow o ft en do exp os ur es occ ur ?) and tim.in g (at what age do ex po sures occur ?). The calculation of actual ex po - sure al so requir es co mpl ex detecti ve wo rk to di scover all kinds of details, in cl udin g the chemical identity (f or ex- ampl e, th e p es ticid e chl o rp y rif os ), source (nea rb y agricultural use), medi- um o f transport (ground water) and ro ute (drinking contaminated we ll wa- ter ). Sc ientists must cons id er this info r- ma ti on on ex po sure aga in st the b ac k- gro und of p eo pl e's activity patt erns, eating a nd drink in g habits, a nd lifestyl e, and they mu st also evaluate the influ- ence of o th er d1 emicals in the a ir , water, beverages, food, du st and so il . Ove raU , this is a daunting challenge. Histor ica ll y, th ose scientists who un- de rt oo k such a co mpl ex task h ave re- li ed on indir ec t me th od s: qu es t io n- na ir es, dia ri es, intervi ews, central iz ed monitoring of co mn1ul uty a ir or wa ter, and a reco rd of br oad activity pa tt e rn s among the population. But th e results were o ft en di s app ointing. Although th ese ci rc um stantial approach es ha ve the adva nt ages of pr acticality and fru- ga li ty, th ey can also intr od uce substan- ti al unce rt a in ty into r es ulting exposure estimates. Tltis sho rt co ming muJtipli es th e potential for a fundamental e rror - cla ss ify in g a person as "not ex po se d" wh en he or she has b ee n or vice ve rsa. A seco nd a ppr oach, the direct mea- surement of an individual's environ- me nt , is so metimes a possibili ty- for ex- ampl e, a person might ca rry a porta bl e monitor to reco rd co nta ct wi th a ir bo rn e che mi ca l s. Al though this technique of- fers an unequivoc al record of chelnic al contact, it is tec hn ol og i ca ll y infeasible or p ro hibiti ve ly ex pensive to measure most po llutants this wa y. Also, although sud, monitors do cum ent exposure, they te ll nothing about the pe rson's upt ake o f th ese airbo rn e chemicals- h ow mu ch truly gets i nt o his or her bod y, which is, of co ur se, th e m os t relevant © 2004 Sigma Xi, Th e Scienti fic Research Societ y. Reproduction 38 Ame rican Scientist, Vol ume 92 wi th pe nn ission only. Contact pe nn s@ amsci.org. • '::: \ . . , , • Bettmann / Co r bis Figure 1. In Jul y 1945, DDT was widel y (and mistakenly) hailed as a progressive measure to eradicate di sease-bearing mosquitoes without po s- ing a ri sk to human health. In this photo from a be ac h on Long Island, New York , a new insecticide-s pra yi ng machine is tested as beachgoers play in th e mist. Although this chemical contact is obvious, many other sources of environmental chemical exposure are more difficult to iden- tify. Human biomonitoring exami n es people's blood and urine to evaJuate actual l ev els of more than a hundred substances. information fo r assessing health risk. Fortwlate l y, technologic al ad vances in biomedicine and analytical chemistry now make it possible to ge t exac tl y tills info rm a ti on. Biolllonitoring meaSUTes the actual l eve ls of suspected environ- mental chemicals in hwnan ti ss ues and fluids. This third approac h has come to be ilie "go ld s tanda rd" for assessing ex- posure to chemicals. Blood (and Urine) Will Tell Bio monito rin g is not new. It has its roots in ilie analysis of biological sanlp les for markers for various pharmaceutical compounds and occupational che mi - cal s, efforts that so u gh t to preve nt the harmful accu mul a ti on of dangerous substances. A1tllOugh it h ad a differe nt name at the time, the ge neral idea was www.americanscientis t. org first applied a bout 130 years ago when doctors mo n itor ed the amount of sa li - cy luri c acid in the urine of rhe unl atics who we re being treated with lar ge dos - es of salicy li c acid (ilie precursor of as- pir in ). And as ear ly as t he 1890s, fa ctory wo rkers who were exposed to l ea d h ad the ir bl ood and urine screened to fore- stall ilie elevated levels that pr oduced acute lead poisoni ng. The se in ves ti gators soo n lea rn ed that the degree of contact wi th a sub- st ance d oes n't necessarily determine the biol og ically relevant expos ure to th at ch emical. As a result, th.i s measure didn 't help mu ch in pr ed icting th e risks of l ead poiso nin g. However, they did f ind that the amo unt of a co m- pound tl, at crosses th e bod y's bound- a ri es {ca ll ed the i.nt ernal or absorbed dose , or so metimes ilie body b urd en) has con siderab le val ue for estinl ating the risk to h ea lth . Today, it is relati ve ly aff or d able to m easure the absorbed do ses for hu n dreds of che mi cals by l ook ing for biomarkers of exposure in access ibl e human ti ss ues a nd flui ds, in- cluding saliva, seme n, urine, sp utum, h air, feces, breast milk. and fingernails (all of which can be co ll ected r ead il y), a nd blood, lung tissue, bone marrow, f ollic ul ar fluid, ad ip ose tissue and blood vessels (which require incursion into th e body). Alth oug h pro cedures to co ll ect any of the first se t wo uld , tech- ni caU y, be considered "noninvasive," in fa ct, that categoriza ti on rests on cul- tural, psychological an d social factor s. So obtai nin g the ri ght material can some t imes be aw kward. Fortunately 2004 January-February 39 • • • •• oXlcan toxicant 2 exposure assessment emission source l • • . • • pathway ~ • 1 • potential dose ~ I absorption barrier internal dose adverse effect • Figure 2. Which toxicant is morc dangerous? Because of th e mu ltiple st ep s through which an en- vironmental chemical mu st pass before it become s a po tential health threat, th e answe r is not al- way s clear. Here, toxicant 1 is more abundant in the e nviron m en t, but th e specific properti es of the chemical may mean th at it pos es less medical risk than ano th er compound. Different methods of exposure assessment can eva luate each of these steps, but biomarker analysis, which measur es in- ternal do ses of specific s ub stances, provides the m os t relevant information for human health. for th ose of us in th e biomonitoring for the pr esence of bi ological markers fie ld, it's never necessary to collect all of exposure-genera ll y the targeted of those samples- blood a nd urine are chemical, it s primary me tab olites or typica ll y sufficient. These are analyzed the pr oducts of it s reac ti on wi th certain 40 Ameri ca n Scie nt ist, Volume 92 natural co mpound s in the bod y, such as pro tei ns. Choosing the appropriat e ti ssue or fluid fo r biological monitoring is based pr imarily on the chemical and physic al pr operties of the chemical of interest and, in so me cases, th e time interval s in ce th e last ex po s ur e. For exampl e, so me chemicals including diox in s, polychlorinated biphenyls a nd organ o- chlorine pes ti ci des have long biological res id ence times in the body (months or years) becau se they are sequestered in fatty ti ssu es. They are thu s sa id to be fat-loving or , to use the pr oper term, lipophili c. By co ntr ast, ot h er chemicals such as orga no pho s phat e pesticides and volatile organic co mpound s, which don't acc umulate in fats (being li po- phobi c), have relatively short biologi cal residence times (ho ur s or days) and tend to be metabolized rapidly and ex- cre ted in the urin e. Th e tim e s in ce th e l as t exposure can also pl aya key role in detemlining the best biological spec im en for analysis. For example, a persistent chemical, such as a dioxin, re mains present in blood for a much longer period (years) th an do es a nonpersiste nt compound such as ben- zene (hours), but dioxin does not form Signifi ca nt urinary metabolit es, whereas benzene does. For these reasons, persis- tent chemicals are typically measured in blood, and nonpersistent che mi ca ls are measured in urine (as soon a ft er expo- sure as poss ibl e), although th ey can also be detected in blood soon after ex po- su re if th e analy ti ca l methods are s uffi - ciently sens iti ve- and they usually are. Specia li sts ca n n ow detect ex tr emely l ow levels parts-per-billion, parts-per- trillion, even parts-per-quadrillion -<lf multi ple markers us in g a rela ti vely sma ll sample, say, 10 milliliters or less. Clea rl y, the sensitivity of the analysis is important in chOOSi ng what to mea- sur e- but it's not everythin g. Other is- sues must be conside red before th e re- s ult s can be consider ed meanin gf ul. We ll before attempting to discern trace amounts of target chemicals, an inves- ti gator should be ab le to answer three br oad questio ns: How is the meas ur e- ment related to the magnitud e, dura - tion, frequency and timing of expo- s ur e? H ow do sub se que nt processes within the body-;;uch as absorption, distribution, metabo li sm and excre- ti on- influence th e targeted biomark- er? And is this particular marker spe- cific for a ce rt a in che mi ca l or d oes it ind ka te an enti re cla ss of subs tan ces? Because the science underpinning human biomonitoring has improved significantly in rece nt years, these qu es- tions are n ow easier to answer. The rap id adva nc ernent in knowledge of wh at the body does to chemicals that are inhaled , ingested or absorbed through the skin has led to better inter- pretation of the range of concentrations for various biomarkers. And the num- b er of t estable compounds h as in- creased dramatically: Sensitive and spe- cific bi oma rk ers are ava ilab le f or man y envi ro nmental che mi cals, including metals, dioxins, furans, po lychlorinated bip henyls, pesticides, volatile organic co mpounds , phthalates, ph ytoestrogens and environmental tobacco smoke. As research continues, the li st will surely continue to grow. Exposure mId Uptake Bi omo nito rin g has many advantages over traditional methods. For example, biological s.unp les reveal the integrated effects of repeated co nt act. Also, this approach documents a ll routes of expo- s ur e- inhalation, ab so rption thr ough the sk in a nd ingestion, including hand- to-mouth tra nsfer by c hildr en. Such spec imens also re fl ect the modifying in- fluences of physiology, bioavailability and bioaccumulation, which can mag- nify the concentrations of so me envi- ronmental chemicals enough to raise them above the detection threshold . Perh aps most importantly , these tests can help es tablish correla ti ons b etwee n expos ur e and subseque nt illness in in - dividua l s-w hich is o ft en the key ob- servation in proving whether or not a link exists. A great strength of biomonitoring is th at it pro vides Wlequivocal evidence that both exposure a nd uptake have tak- en place. In some cases these da ta can confinn the findin gs of traditional expo- su re estirnates. For exam pl e, in 1 979, 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 conce rnin g because many peo- ple in that area cau gh t and ate the fi sh regularly. In response to this discovery, th e Centers for Di sease Control and Prevention (CDC) constructed an eval- uation based on DDT concentrations in fish a nd the amoun t of fish eaten per week. This estimate indeed correlated with levels of DDT and its metabolites, www.ame ri canscientist.org food soi V dust water levels levels levels air nu:;~:~~al levelS j/ heallh h leslyle ~mathematlcal predicted level of toxicant personal • modeling + in people hag::etic ~t ~ predisposition /~ lung , intestine and' i excretion skin absorption rates metabolism accumulation j Figure 3. Traditional esti mat es of human exposure have to account for many variables, in· eluding so me that demand assumptions about factors that are poorly understood. Th e resu lt is often uncertain. human tissues or fluids personal environment or microenvironment ~ emission ~ accuracy Figure 4. Exposure to environmental chemi ca ls c an be assessed in several ways. Generall y, th e accuracy and cost vary togeth er. Monitoring emission sources is the least expensive and least accurate means of determining human exposure, whereas biomarker meas ur eme nt is mo re costly but also highly informative for that person. 2004 January-February 41 Figure 5. At its Environmental Health Laboratory, CDC scientists use several types of high- resolution mass spectrometry to analyze human tissue and fluid samples. The equipment shown here is being used to measure dioxin levels in a sample of blood se rum. (Photograph courtesy of James L. Pirkle.) DOE and DOD, in the blood of Triana res id ents. In a similar story that un- folded in U, e late 1980s, chemical-plant workers in New Jersey and Missouri discovened that they had been exposed to dioxin-contaminated compou nd s up to U, e ear ly 1970s. They h ad come into contact wi th th e di ox in in various ways- breathing it , swa ll ow ing it and taking it in through the skin. Despite the comp lexi ti es of th e ir interaction w ith this dangerous sub stance-a nd the time interval s in ce expos ur e-a scheme that used occupational records to calculate the duration of potential expos ur e was able to acc ur ately es ti - mate inte rnal doses. This finding was confirmed by the correlation of these results with th e conce ntr ation of djo x- ins in th eir blood. Hav in g information abo ut expos ur e alld uptak e is more than a pro forma de- tail: There are many cases in which tra - ditional estimates of exposure (q ues- tionnaires, proximity to so urces, environmental concen tr ations, con- structed scenarios) are not correlated with measured biomarkers. For exam- ple, from 1962 to 1971 , the U.s. Air Force sprayed the defoliant known as "Agent Orange" in Vietnam. Many ser- vice members who participated in that operat ion touched or breathed the her- bicide, potentially exposing themselves 42 Ameri ca n Sc ientist, Volume 92 to hi gh levels of dioxin. The Air Force fir st estinlated th e risk to soldiers using a sce na rio approach, which included the average dioxin concentra ti on in Agent Orange, the number of ga ll ons used during a so ldi er's tour of duty , and the frequency and duration of p0- tential con ta ct based on job descriptiOll. Despite a co nsiderable sc ientific effort that went into these pr edic ti ons, CDC studies in the late 1980s proved that none of th e exposure estimates were corre lated with the measured blood levels of dioxin in at-risk troops. A sub- sequent investigation of personnel with th e hi ghest dioxin levels did iden- tify so me patterns that explained their in creased con tact - for exam pl e, small- statu red enlisted men o ft en climbed into the chenlical tanks to clean out residual Agent Orange. A more st riking examp le of the val- ue of biomonitoring came in the mid- 1970s when the United States elected to start phasing o ut l eaded gasolin e. Prior to this decision, traditional mod- els had suggest ed that eliminating lead in gaso lin e would have only a s li ght ef- fect on people 's uptak e of th at metal. However, biomonitoring data from the CDC's Second atio nal Health and Nutrition Examination Survey re- vealed th at from 1976 to 1980 (as un- l eade d fu el was first introduced and gasoline lead decreased by approxi- mately 55 percent) there was a parallel decline in the amolmt of lead coursing thr ough th e veins of the U.S. popula- ti on. Overall, average blood concentr a- ti ons decreased from abo ut 16 to less th an 10 micrograms of lead per deciliter of blood. These data demon- s tr ated the effectiveness of removing lead from gasoline, and th ey were a dominant factor in the decision by the Environmental Protection Agency (EPA) to remove l ead from gasoline m ore rapidly -a ta sk th at was effec- tively complete by 1 991. Today, the av- erage blood-lead level in the U.S. pop- ulation is less than 2 micrograms per deciliter Exposure Disclosure The st ud y th at revealed the tight co n- nection between U, e lead in people's gas tanks and the lead in th eir blood was mounted by the CDC, which conducts the Na ti onal Health and Nutrition Ex- amination Surveys (NHANES for short). A lth ough no environmental chemicals were measured as part of NHANES I (1971-1975), starting with NHANES II (1976- 19 80), the CDC be- gan measuring bl ood lead levels in the U.s. population, ironically enough, af- t er the Food and Drug Administration voiced concerns about possible expo- sures from eating food stored in lead- soldered callS, which turned out to be a very minor risk com paned with leaded gasolin e. As part of NHANES II , the EPA tested for certain persistent pesti- cides in people's blood and nonpersis- tent pesticides or their metabolites in urine. After an eight-year hiatus, NHANES III was conducted in two three-year phases from 1988 to 1994. Tn th at iteration, th e COC measured lead and cadmium and began testing for co- tinine, th e major metabolite of nicotine, in blood. Additionall y, U ,e CDC began a separate pilot program to measure new compolmds, testing for trace amounts of 32 volatile organic chemicals in blood and 12 pesticides or their metabolites in urine from approximately 1,000 of the NHANES III participants. Then came ano th er long gap in cov- e ra ge. But thankfully, in 1999 , NHANES became a continuous survey of the non institutiona li zed U.s. popu lation. (It is thought that excluding members of isolated organizations, such as mili- tary personnel, college students and prisoners, provides a better cross-sec- ti on of America.) In th e cu rr e nt design, Identifying priority exposures. Out of thousands of chemicals, whi ch are the most dangerous? Biomarkers can help set priorities for public health and regulatory C fo ll ow-up. 1 Identifying at-risk populations. Large biomarker studies can distinguish exposure differences among racial, geographic or socioeconomic groups. Providing integrated dose measurements. Biomarker analysis provides a direct assay of body burden that integrates exposure from a ll sources, even ones that are hard Recognizing time trends in exposure. Periodic measurement of biomarkers in the population shows how body burdens of chemicals vary from season to season, year to year and decade to decade. • Establishing reference ranges for comparison. A blood test shows that you' ve been exposed to some chemical. Should you be worried? Your doctor can't tell without data from people with little or no exposure. Evaluating exposure prevention efforts. Our government is entrusted with reducing people's exposure to environmental chemical s. Do th ey succeed? Before-and-after biomarker tests can tell. 1 to measure. Fig ur e 6. Wh en used to establish levels of human chemical exposure, biomonitoring has six major u ses that ca n help to pr otect public hea lth. www.americanscientist.org 2004 January-February 43 110 "' c Jl 100 '0 "0 '" 90 c '" 0 ~ '" 80 E- w 70 :a C '" 60 0> '" . ~ 50 "0 w '" ~ 40 "0 .!!! '" 30 Predicted blood lead ::::r 70 17 :g, 16 "t;; 60 c 15 ::; "0 50 1 14 :9 o; '" j 40 - 13 > .!!! g> 12 "0 ~ 30 8 :0 ~ 11 c '" ] '" 20 " E 10 "0 10 9 .!!! '" ~ 0 - I , I 1975 1976 1977 1978 1979 1980 1981 1965 1970 1975 1980 1985 1990 1995 y ear year Figure 7. Leaded gasol in 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 th e changes in gasoline (left). Blood le ad and gas lead continued to fo ll ow nearly identical decreases up to 1990. At the same time, a series of studies on le ad toxicity showed that lower doses could still cause adverse effects, prompting a steady decline in the level defining lead poisoning (right). a new national s ampl e is coll ected every two years. Although so me other studies have locused on sp ec ifi c popu - la ti ons or on more r es tricted dat a, NHANES is the only na ti onal s ur vey that includes both a medical examina- ti on and collection of biological sam- ples from participant s. In dividuals se- lected for NHANES are representati ve of the U. S. population, me aning that th ey do not necessa ril y have high or unu sual exposures. About 5, 000 parti c- ipants are examined annually fr om 15 loca ti ons thr oughout the co untr y. Reporting For Duty In March 2001, the CDC released the National Report on Human Exposure to Environmental Chemicals, which in- cl uded data from 1 999 on 27 chemicals. A seco nd report wa s published in Jan- ua ry 2003 that examined 11 6 chemicals in sa mpl es fr om 19 99- 2000. Both stud - ies used biomonitoring to provide an on go ing assessme nt of exposure to a variety 01 sub stances. Although vari- ous studies of workplace ex po s ur e, lor example, had rai se d concerns about the he alth e ff ects of such chemicals, most of them had n eve r before been measured in a re pr esentative slice of the U.s. population. The inventory of tested substances in th e second CDC report includes lead, mercur y, cadmium and other metals; persiste nt (organoc hl orine-based) and no np ersiste nt (o rg an ophos phat e- and ca rbamat e-ba se d) insec ti cides, herbi- cides and other pesticides; pest repel- lents and disinfectan ts; cotinin e; phtha- lates; polycyclic aromatic hydrocarbons; di ox in s, furans and pol yc hlorinated bipheny ls ; and phytoestro ge ns. Results fr om the general population are subdi- vided by ag e, gender and etluticity. An important fe a tur e of the CDC re- port is that it provides reference ranges for ex po s ure amon g the general U.S. population. If peopl e ar e c on c erned that they may ha ve be en exce ss ively exposed to an environnlental chemical, th ey ca n compare their bi oma rk er lev- els to those standard s. These reference ran ges a re immen se ly beneficial to publi c- health s ci entists who mu st de- cide il c ertain high-expos ur e g roup s need foU ow- up action. If average levels among the cohort are similar to those of th e general publi c, then the group 's exp os ur e is unlikely to cau se unique problems. On the other ha nd , if levels are s ub stantiaUy hi gher than na ti onal Figure 8. One important function of biomon· itoring is that it can identify specific subpop- ul ations that may be more vulnerable to ex· posure from a particular chemical. For example, p,p'-DDE, a long-lasting metabolite of DOT , is more than twice as high in Mexi· can-Americans compared with the general population. By contrast, cotinine levels are the l owest among this group, indicating that they have th e least exposure to environmen- tal tobacco smoke. For both cotinine and lead, non-hispanic blacks s how ed the highest lev- els. DOE (in na no grams per gram of lipid) and l ead (in micrograms per deciliter of blood serum) data are from th e CDC's Sec- ond National Report on Human Exposure to Environmental Chemicals, published in 2003. Information on cotinine (in nanograms per milliliter of blood) is from the third National Health and Nutrition Examination Survey (NHANES III), 1988-1991. no rm s, epidemiologists can confirm th e unu s ual ex po s ure , ide ntif y th e so urces and provide continuing health care as appropriate. The reference ran ges pr ov ide indirect fin ancial ad- 700 ai :g: 600 We ~ u:; 500 . ;:: CI Cii CI 400 EE g 2 300 0> :Jl 2 00 1 00 .L -' ' ' '- -= ~_ 0.5 0.4 0.3 0.2 0.1 -' ' ' ' ' -'- - 3.0 44 Ameri can Scie ntist, Vo lume 92 5 4 nonsmokers smokers 0.1 1.0 10 100 serum cotinine (ng/mL) 1000 vantages too, bec au se distinguishing common from unu sual chem ic al con- tact helps dir ect reso ur ces to the most- pertinent ex po s ur e situations. The overarching purp ose of these re- p or ts is to help scientists, ph ysicians and health o ffi cials to prevent, reduce and treat envir Ol IDl enta ll y induced ill - nesses. However, some caution must be exercised in interp re ting the finding s: It is important to remember that detect- in g a chemi ca l in a per so n's bl ood or urine does not by itse lf mean th at the ex posure c au ses disease. Se par ate s ci- entific studies in animals and hum an s are required to determine w h.i ch levels are likely to do h ar m. For most che nu - cals, tox icol og ists simply don't ha ve this info rm a ti on. But even if scientists are not s ur e of the overa ll level of ri sk, they can make co ncrete statements about whether sit- ua ti ons are getting better or wors e. The la test CDC report, in addition to listing c urr ent biomarker levels in the popula- tion, al so hi g hl ights so me interesting expos ur e tr e nd s gleaned fr om earlier N HA NES findings. For example, from 1 99 1 to 1 99 4,4.4 percent of children be- tween the ages of one and fi ve had lev- els of bl ood le ad gre at er than or equaJ to 10 micrograms per deciliter, the Fed- eral ac ti on leve l. By the second co ll ec- tion period (1999 and 2000), o nl y 2.2 pe rcent of this age gro up exceeded this t hr eshold . Ti m decrease suggests that eff or ts to reduce lead expos ur e for chil- d re n have b ee n successful. It also serves as a reminder that so me children, in - cl ud ing th ose li ving in homes with lead-based paint or lea d- contaminated du st, remain at un acceptably hi gh risk. The last r epo rt al so indicates a hope- ful trend in the expos ur e to environ- mental tobacco smoke, as sh ow n by tests for th e bi omarker cotinine in the bl ood of nonsmoker s. Median lev eJs of cotinine fell more than 70 pe rce nt in roug hl y a decad e- that is, between th e second (1988 to 1991) and third (1 999 and 2000) periods of dat a co ll ec ti on. T hi s dr op provides o bj ec ti ve ev idence of reduced exposure to environmental tobacco smoke f or th e general U.s. pop- ul a ti on. Nevertheless, th e fact that more than half of American youth continue to be exposed to environmental tobacco smoke re mains a publi c- health co ncern. The CDC plans to release future re- ports that document their biomonitor- ing e ff orts every t wo year s. In the next edition, they w ill also a dd the findings from separate studies of special popu - www . ame ricanscientist.org Figure 9. U.S . population cl early segregates into smokers and nonsmokers based on the level of c otinine in blood. The working th.reshold for distingui s hing the two groups is ]O nanograms per milliliter of blood serum . Among nonsmok ers, the highest values of cotinine were found in children under 12 , and they were strongly reflective of the number of smokers in the home. Th e data are from NHANES JII , 1988-199]. lations, such as the laborers who apply pes ti cides to crop s, people li ving near hazardous-waste sites and workers in le ad smelters, all of which a re likely to have hi gher-than-average ex pos ur es to certain enviro nm ental chemical s. Annual Check-Up With Biomarkers? As th e 21st ce ntur y unf olds, the C DC s ur veys and other we ll -designed bi o- monito ring s tudi es will co ntinu e to build an under standing of peopl e's e x- posure to tox ic environmental chemi- cals. Nonetheless, these dat a will not obviate the need to co ll ec t o th er kinds of rel eva nt informa ti on- to monitor so ur ces of pollution, to co ndu ct s ur - veys of tox ic s ub stances in the environ- me nt and to s tud y human ac ti vities and behaviors that contribute to expo- sure. Moreov er , further research in to x- i co logy and epidenuology is necessary before specialists ca n int e rpr et th e h ea lth Signific an ce of ex p os ur e bio- markers for most environmental chem- i ca ls. Particularly as detec ti on methods impr ove- en abling inves ti g ator s to rn easure lower concentrations of more chem ic al s from sma ll er s amp les at less cost -scie ntific und erstanding of w hat th e body does to th e chenu ca l ( and vice versa) mu st keep pace. If this e ff ort is successful, a full scr ee n of expos ure biomarkers may be a part of eve ry ro utine ph ys ical ex am in the not- too-distant futur e. Bibliography DeCaprio, A. P. 1997. Biomarkers: coming of age for environmental healt h and risk as- sessmen t. Ell v irolllll e lltal Sc ie ll ce & Te e/molo- gy 31 :1837-1848. Mendelsohn, M. L. , j. P. Peeters and M. J. Nor- mand y, eds. 1995. Bi omar ke rs alld Occ upa- tio ll al Healtll: P rogress an d P erspec ti ves. Was h- ington, IX : joseph He n ry Press. Me nd e lsohn , M. L. , L. C. Mohr and j. P. Peeters, eds. 1 99 8. B io mark er s: Med ic al and Work pla ce Applicatio ll s. Wa shington, DC: joseph Henry Press. Needh am , L. L. , and K. Sexton. 20C10 . Assessing ch il dren 's exposure to hazardous environ- mental chemicals: An overview of select ed research cha ll enges and c omp lexities. /our- lIal of Expos ure Analysis alld Etf v iroll1" elltal Epid emiol og y 10 (Pa rt 2) :6 11-629. Ne edham , L. L. , D. C. Patter so n, Jr., V. W. Burse, D. C. Paschal, W. E. T urner and R. H. Hill, Jr. 1996. Reference range data for as- sessing exposure to selected environmental toxicants. To xicology alld Illdus tria l Heaflll 12 :507- 513. Pirkl e, J. L. , E. J. Sampson, L. L. Need h am , D. G. Patterson, Jr ., and O. L. Ashle y. 1 995 . Using biological mon itoring to assess human expo- sure to priority toxicants. Ell v irollm elltal H ea ltll P erspec ti ves 103 (supplement 3): 45-48. For relevant Web links, consult this issue of America" Sc ie llt is t O l/Iill e: http ·//w ww ametican SC ientjs! m:g I I SS u eTOC/ i ss ye / S21 2004 janu ary- February 45 . Human Biomonitoring of Environmental Chemicals Measuring chemicals in human tissues is the "gold standard" for. method of accurately measuring not only co ntact with, but al so ab so rption of toxic chemicals fr om, th e environ- ment- human biomonito rin g. Is It in Me? Each person's risk of. other sources of environmental chemical exposure are more difficult to iden- tify. Human biomonitoring exami n es people's blood and urine to evaJuate actual l ev els of more than