Tobacco Documents Online

SENT BY:Xerox Teleccpier 9021 +12-14-93 +_10.44 l I , -4 49 2203 303362 „ 2 ..:--o. _....,, , Vie111f '~'."y'Ff ' . a.. ~~~ = Acfual Causes of Death in the United States J. Michael McGinnis, MD, MPP, Wiiliam H. Foege, MC), MPH .ObJecfilvc.-To Identify and qua.rztify the major external(nonger,etic) factors that oarltribute to death In the United States. Data Sources.--Artic-ies published between 1977 and 1993 were Identifie-d through MEDUNE searthes, reference crt-ations, and expert consuttation. Govem- m*nt reports and camp3fations of vltal statistics and surveiUance data were also oi}- t2ined. Study Seleation.-Sources selected were those that were often clted and those tha.t ir6cated a quantitat~ve assessment of the relatfve confributlons of varfous fac- tocs to mortality and morbidty. Data Extractlon.--0ata used were those for vrf)ich spEcific methodoogical as- scsrr4;tlon.s were stated. A table quantifying the contributions of leadng factors was corztr=ed using actual c9unts, gerterally ar:cepted estimatss, and cslcutated es- t3zttates that were developed by summing various indfvidual estimates and correct- Ing to avoid doubie counting. For the factors of greatest cAmplexity and uncertainty (d;et and acevity+ patterns and taxlc agents), a consarvat)ve approach was taken by dxasing the lower boundaries of the various estimates. Data Synthesls.-"fbe mast prominent corstn-butcrs to mortaTity In the United Siata:s in 1990 were tobac:,o (an estimsted 4Dp G00 deaths), dJet and dctfvfity pat- tems (300 400), alcohol (100 000), microbial agents (90 000), taxic agents (60 000), flrear[ns (35 000), sexual behavior (30 000), motor vehicles (?5 000), and illictt use of d.vgs (20 000). Sao;oecanomic statvs and access to n,edical care are also im- paZant ooninbutors, but dNcuit to quantIfy independent of the other factors cited. Bec:atsss the studies reviewed used drtferent approaches to derfve estimates, thd srated numtaQra should be viewed as ilrsi 8ppt~mations. Conclusions,--Approtmateiy half of all de2ts that occ:urrad in 1990 could be aitrExlted to the factors identified. Atthough no attempt was made to further quantify ttze irrspact of thesa factors on morbidrty and quafrty of llfe, the pub8c hea,lih burden tthey impose Is considerable and offers guidance forsha.ping heaSth policy priorities. WAx.L iV9&9702W7-s= IN 19W, apprwdmataly 2 148 000 US reaideats died. C4a-dfeatr.a filed at the t3tne of death indiesie that t.hea deaths were moAt cornmon.l,y due to heart dis- Qsse (720 000), cancer (505 000), cg.re- brovascalar disease (144 000), accidents (920C)0), chronic ob~ctzve pulmonary disPSSe (87000), pneumotia and influ- eaza (80 000), diabetes raellitaa (~',S 000), suicde (a1000), chronic liver disease and dzrho:dz (2b 000), and hwnan irarnuno- deftcencyvtrus CAITr}infection (25 000)? 0.'tsaraferenoed as the 10 leading tisses of death in the United 8tates, they gen- erzlty indieste the prsmary pathophysi- Fsom h US D.para7,ent d t4a.M r,a Hurn~ srr- *~a~~ DC (Dr jd.V3rxsa), er4 Vr® Cart.r Prsri~re.+ cars.r, Ati+ns, c3t (Or r-p.3. Awaprlre na.•n t5 e+e D." Aa&tmau s.e,awy tar Fiost% owe„ prv.+rxcn W,a Ke.rp, PraTOOOrti us Deoartr~re oc t&.en &-d rirTon sWACs s3o c sc Sw, ftrn t1SZ 4+aaM-qxra DC aC26t cDr f cc3s+r+a} otogic3l conditions identifed at the time of death, as opposed td thdr rrrot Guses, ' These conditions actuaIIy result from a cambinstion o,f inborn (largely genetic) and external fsctbzla. Berssnae most diaeases or ir~7uries are mIIltifacfrarisl in ns.ture, a key challenge is sorting ont the ralative contributions of the various factnra. For heazt die- ease,'.vaIl-cstablished external risk fa-^- tors include tobacco use, elevated se- rum cholesteral Ievela, }iypertenaioa, obea;ty, and decrra.sed phy$ical activ- ity; for verious caneesa, such risk fa.o- tors include tobacco ase, dtetary pat- terna, certain infectbovs agents, and ett- viron.•nental or occupattonal Psposure to earcinogenic agenta. Even motcir ve- hic]e try uries can be associated with mu2- t3ple facrtors, including alcohol uae, fai7- u.r® to use passengRr proteclaon systems, poor rcadway design, and inB.dequnte Iaw enforcement'hhese fmctarz raay act independently of each other, the risks being additive according to the effect of each, or they rnay act apnergisl3cail,y, the inttraction of factors presenting a greater total risk than the sum of their individua2 effe.&s. _ Ava2able analyses of the roles of vari- cus external factors in these coaditions suggest that the mob t prorainent idan- tMabla cantrib7utcus to death among US residents are tobar.co, diet and act3t~Cty patt2zas, alcohol, mia-obial agents, taadc agents, Ere9*mr, sexus.l behavior, mo- tor vehidea, and Micit use of dnugs, When thare contnbuta to deaths, those deatha see by defSnition pretnEtnre and are of- ten preceded by impaired quality of life. Althoagh mortality is but one measure of the health status o.f a nation, th.e pub- Lic hwzith burden imposed by these con- tr0cutors offers both a mandate and guid- an ce for shapfng h ea.ith policy priorlties. M1=T'1-IODS 'I'bi,s artide 'summar-ses published re- pnt.s that athzbutad dest'), to these con- tzioutora and presents a composite ap- Frazimation nf the totals repazted for earh ('I`abl e). Artzcl ea published be twesn 1977 and 1993 were ideat!:3ed through Il1EDLINE searches, ze'eiCnoe dtations, and expert cansnl `~ation Governroent re- ports and eoarpzations of vital statis`aca ar,d surveillance detfl were also obtained. All rekvant ansl,yses srere reviewed In fv.Il. Tbose selected for use in devrlop.ing estfm~ ware those mogt oltan dtnd and those for which the methodologieal assumptions could be identiied. The Iimitatzlons in the data should be underscored both with respect to daS- denaes in the primary datsbases (eg, the psudty of data oa the role of drugs in mot-0r vehicle htalltiea or on long- term exposure levels, of populations to various toxic Lgents) and to the dispar- ate approaches used in the studiee zr- viQ W ed to arrive at estimatas of th e can- tnb ut3 on of a faetor to n puticZSlar h eslth outeome. In some cases, assignmenta were aitempted through simple tallieB of availe infarmation about the presence or absence of a facto.r in assocdation with a gjveSs 6utzam8 (eg, whether or not a .l9Ja,A. Nwwrbw 10, 19S3-Va' 270, Ne.. 18 Acax! Gauses d De,ac'i h tTie lJ-tea StaiaS-'.4cGinnie & F^.,e9e 2207

StNI dY:Xe ox elecopier '1u11 ~1Y-14-9~ I 1u.45 I -j c~ 11u3 3u~3b1 „ 3 driver in a motor vehicle fatality had a bleod aleohol coneentration above a cer, Wn level). In other casea, populatson- a,ttributable risk calculations were used to arrive at estimates based on determi- nations of the relative risk for a particu- lar health outcome of a population ex- posed to a spedfed health risk Soine of the studies preaented meta-analyses of sepozts in the litsrature on a given tapic. Estimates were often liznited by the ad- equacy of infornzat; on as to di_,eace preva- lence, risk factor prevalence, and the na- tare of the relationship to other contril}- uting risk factars for the disease. Despite their limitat3ons, the results of such studjes provide a sense of the relative impact of various fa.ctors on health in the United States. Derivation of the numbezs presented in the Table ia expls.ined below in the discussioa of eanc category. Where well-established methodologies have been developed for msking the estixnates, as with tobarco aad alcohol, they have been used ap- proxun,ately as reported. For areas of greater uncertainty, such as diet and activity patterns and tos,ic agents, a sum of the lower boundaries of the estimates for vsrious disa>sse Quteomes has been used. Although several of these factors .are interrelated In their actions, care has been taken to avoid double count- i;ag. Given the fragility of the database involved and the fact that the studies cited use different appraaches to derive es'dm.ates; these numbers should be viewed as f[rst apprcximations. and 20% to 80% of Iow-bir~wrignt in- fznta.lL' APproXJTnstely 30Q0 lung csn- car deaths annually among nan4nokers have been attributed to environmental tobacoo smake.a The sum of the lower and upper boundaries, r,A-pe,:tively, for these estimates would yield an approxi- mate range of 257 000 to 4F$ 000 tobacxo- sttributable deaths in 1990. Using a spe- cially developed sofl-ware package,a the Ce•nta-e for Disease Contml and Preven- tfon (CDC) estimated that 418 690 deaths were caused by tobacco in 1990, including appraximataly 30% of all caneer deaths and 21% of cardiovascuiar disease d eatha'TS' The CD C estimates have been widely arcepted and provide the basis for the 400000 ffgure included in the'I`a.bla. Diet and Activity Patterns Dietary factors and activity pattezrs that are too sedentary are together ac- countsble for at least 800 000 deaths each year. Dietary factors have been asa.ad- ated with rdrdiovsscular disaaasFSs (rore- nary arte.ry disease, sfroke, and high blood pres.aure), cancers (colon, breast, and prostate), and dia.betes mellitus u Physical inactivity has been aswdated with an incrraaed risk of death for heart disease'L21~-- and colon cance.r." The in- terdependence of dietary fa-hrns and ac- tivity patterns as risk factora for certain diseases is Muatrateti by the case of obe- sity, which is assoeated with increased risk for cardiovascular disease, eertain cancers, and diabetes, and is dearly re- lated to the ba]aace betwee.n cslories eon- s-:med and calories expended through metabolic and physical activity. Sim17s.-ly, high blood pressure, a major risk for sl7oke, can be affected by dietary so- dium, obesity, and sedentary Iifestyie. The Carter Center review of deaths in 1980 atributed 290000 deaths to over- rrutrition and another 297 000 to high blood pressure.` Sedentary lifestyles have b$en linked to 23% of deaths from the leading chronic diseases." An ass2ssment of the decline in enronary artery disease mortality from 1.968 to 1976 credits rr- ductione in serun cholesterol levels with about a third of the improyement." Some studies eredit changes in sodium con- sumption ulth the ~otzntial to lower death rates for coronary heart dieasse by 16% to 30% and stroke death rates E8% to b5%845~ Hs1f clall type II dta- bates (non-insulin-dependent diabetes mellitus) is est3zaated to be preventable by obesity coatroLu A 50% reduction in consumption of arumal fats might result in a proportionate reduction in risk for colon csncer?' In the most extensive analysis to date of studies on risk fa.ctozs for cancarr, Doll and Peto established 35% as their best estimate fnr the proportion of all cancer deaths attibutable to dietLu Aebua! Causas of Dva3h fn !hs Un}lad Statas in 1 e90 r a..4ss ~ o! rt '~ RESULTS Tobacto Tobacco' accounts for approximately 400000 deaths each year among Ainari- carma. It contnbutes substantiaIl•y to deaths fzom cancer (especi.al}y cancers of the lung, esophs.gus, oral cavity, pancrz.ae5 kriney, and bladder, and perhape of other organs), cardiovascular disease (enronaTy artery disease, stroke, and high blood pres.sure), lunQ disesse (chronic obetavctive pulmo- ns.ry disease and pneumonia), low birth weig}it and other probleras of infancy, and burrL--= In a ma,j or e~'Oxt that drew on analyaes that had been coznmissioned to assess the mortality, morbidity, and $- nabrSal burden imposed by each of 15 pri- ority health problems' the Carter Cen- t.e3's Closfng the Gap project attributed 17% (338000) of all deaths in 1980 and 1346 of all potentiai years of life Iost frorn' de4th before 155 yw-rs of age to tobacco! Other estimates have placed tobacco's contribution in the range af 11% to 3056 of cancer deathssw 17% to 30% of car- d33ov&cular deaths,Tj03t 8096 of lung dis- ease deatha,"," 24% of pneumonia and in- fluenza death.s,'T 10% of ia..~ant deaths,';" Guae Fio. Totzt 7obwoo a00 000 19 C4.th=v y panoms sao oeo 14 /JCCAci toD = S Mcrobia! agsns so Cr.C 4 Tmdo agerta 60 0?0 3 F17. g ss om Z S.n$1 betmvior SO C"J7 t WA=r ,ahaes Zs so t m=c use or an,es 20 eoo <t Taml • 1 060 000 50 uaa a~dt~ ro,ctus~dse sm ~a.t~rsrg~q tr.m b=Uai uwrsts (eq, fssarrrV) iD pcputatSon a afD- URSDie risY catutLdors (a9, tob&xs). NLms,'•-era ar.r 100 oeo nxe,aw ao d,e n" rer t oa ooc, owf so ow. rou^ded to 4* r"reS 10 00011 be(ov+ so ooo, rwnded to the netreV bobl Other studies have aswc.iated dietary fa.~_ tors or sedentary lifestyles with 2256 to 80% of eardiovascvlar deatha,7•1°~1i_1' 209'0 to 60% of fatal csnces%F10'l and 50% to 80% of diabetes mE1litus cssas,lxx in- duding 3056 of diabetes desths z,1o If the boundaries o.f these various estimates were summed, they would yield a range of apprmdmateIy 809 000 to 582 000 deaths in 1990 related to diet and actv- ity pat#,erns BecausE of the cvmplexity of the issues and the difficulty of the anal,yses relating diet and activity pat- tesns to disaase outoomes, the lower bound is used as the basis for the 300 000 deatha figum presented in the Table- Alcohol Misuse of alcohol acx-ounts for approxi- mately 10.0000 deaths eaeh year, but the related health, sodal, and econonic consequences of alcohol extend far be- yond the mortality tables. An e?titnated 18 mz73ion US residents suffer from al- cohol depsndence,n-" and some 76 rnil- lion are aSected by akahol abuse at some time.1 Estimates of alcohol's death toll ringe from 356 to 10% of deaths?T" Yari- ous estirnates have placed alcohol'S con- tnbution in the range of 60% to 90~'a of cirrhosie deaths,'° 40% to 50% of motor vehicle fataiities,W°A= 16% to 67% of home injurie8, drowziinga, $re fatahties, and job injuries,'-11-" and 896 to 5% of cancer. deatha44-1 The Carter Center project estiznated that 5% of deaths and 15% of potentiai years o.'life lost before age bb were attributable to alcohol use.' $uxnming the boundsries of these esti- xaatee yields an appraxi=te range of 67000 to 107040 alcnhol-related doaths in 1990. The CDC used clinical case stud- ies and analytic epidemiologic studies to detarmine alcohol-attez'butable fractions of various dia.gnose8 aud concluded that a total of 106 095 deaths were caused by alcohol in 1987, including approximately 80 000 deaths from unintentional ir~ju- rias, 19 600 from digestive diaaases in- 4408 .1D,MA Novernber 10. 18S,.'i-vbi 2T0• Nd. 18 Ar_tual ~su&ea o' Dea7i In Lw3 Ureted S7atx---?.4,-,^ .,irrys & F•o"

SENT BY;Xerox Telecopier 7021 ;12-14-93 ; 10:47 ; -0 49 2203 303362;# 4 duding liver cirrhosis, 17700 from in- tantional injuries, and 16000 from can- cers.." Because the CDC eatimate is the one most often reported, it has been applied to 1990 death rates and serves as the basis for the 100 000 aicohol-re•- lated deaths included in the Table. • lL4lcroblaf Agertta Infectious agent,.~--spart from trhose arunted elsewhere with causes of the hu- man L-amuIlodeScienry virus (HIV) in- feation or consequent to use of tobacco, alcohol, or drugs-currentl,y account for appr+oAmataly 90 000 deaths per year. Infeetion8 wese once the lea.ding ldller in the United States, and they are sttll a prominent threat, especdslly to persons ,with otherhealth impairmer,ts. I.•ifeciaous sgents also exert gre.at influence on so- dety through an estiznated 740 million nonfatal zZl.nesses caused by symptom- atic infection.s that occlr annually among Americans!" Although immuriisations and infect;on mntrol snPasurss may a:resdy prevent as many as 185 million iafect.ions and 68 000 deafhs annually in the Unit.ed Staxes," a substential fra.chan of the infeetione and deatha that do occur are aso p:event- abl e. The rnAj or contrlbetoxs to death from tnfectious agents are pneumoeoccal pneu- monia, nxwazrnisl infeedons (in both acute and chronic care faddities), Iegionellosis, Staphyloeoccal au7rcs infeetion, hepati- t3g, and g;oup A streptococcal infecti.ons. Vital statistics repor's for 1980 indirated the number of deaths from infectious and' pmrasitic diseases to be 55 612, phas an- othe.r ?9 518 1~nrn pneumonia gnd 1nIlu- enzs and 7289 fivrn meningitis and en- caghaLris,41 The Carter Center study of deaths ocvining in 19So" estimated that nearlrv200000 deaths cou]d be attributed to infections, of tvhich 13% were pote.n- tislly preventable with current vaccinea." Fiepatitis B infection is a good example. Appraximately 5000 deaths in 1958 re- sulted from heparjtis B infection, includ- Ing about 25% of all deaths from prfmsry liver rancer, although a vaccine has been available since 188Q'~" Tubereslosis, mhich ranked secund as a muse of death in 1900, aa^aunted for 1810 of the infec- tiaus disease deaths in 1990,' and with the spread of antz"biotio-resistant strains, bibereulosi•s gives evidence of increasing in this decade.• The diff culty of assigning responabil- ity fQr infectious dir,c,aae deaths is illus- trated by the fact that, while the number of elasme bacterial pneumonia deaths in- e-easeti about 10% from 19S0 to 1890, those clss.s'u{ed as "othes" and "usspeci- Sed organisin" increased by more than 50% and now account for approximztely 90% of a1 pneumonis deathau A aub- atantial part of the growth in these cat- egor* reflecta the impact of the HIV epide.mic, but most of those deaths are counted in this review under deaths at- tn'butable to unprotected intercouxse or drug use.°° Moreover, many deaths 1~-om pneumonia occar among eanoer, heart, lung, and liver disease patients and are therefore tratsable to other causes such as tobacco, diet, and alcohol (eg, the 24% and 4% of pneumonia and infiuenza deaths aserlbed to tolacta"t and alcohol,13 re- spect:vely). OtheSpneua,onia deaths that may also be related to more prnmmal causes, but which are as yet ursssigned, are counted here as general iafectious disease deaths. Hence, the 80000 deaths included here for microbial agents rep- resent the sum of I990 deaths from key InteratIonsl Cla.ssrfication of Diseases eodes 001 through 189 (infectious and parasitic diseases), 820 through S2B (men- ingiti8 and encephalitis), and 480 through 4ffi (pneumonia and influejsa), and not including those from codes 01.2 thmugh 044 CHIV infect:on) and those otherwise estimated to be attnbutable to tobacco use, alcohol use, sexual behavior, and fI- licit use of druga. Tozlc Agerrfs Estimates of the deaths attributable to toajc agents vary Widely, and because meas-urement techniquss and the recog- nition of health effects are still evolving, the number of 60 000 per year included in the Table may be the most uncertain of the figures indicated for the vari ous o„sesToaic agents may pose a threat to hum1uan~the~s,Ith~1 as occupational hazards, ~Yl Y Liv11:IlGtttsl poly' u t,allts, cont em1TaTts of food and water supplies, and compo- nents of commercial producta They can contribute to conditions that are poten- tially lethal, including cancer and other diseases of the heattt, lungs, liver, lfld- neys, bladder, and neurological system. Estim.ates of the total cancer deaths caused each year by synthetic cbemi.cals in the env-ironment or occupationsl set- tir,gs range up Ward from about 30 000,1-4 including an est.imated 9000 from aabes- tos exposure.' Occupationa.l exposures alone have been estlmated to cause 1% to 856 of all cardiovascular, chronie rea- piratary, renal, and neurological disease death$, as well as all pneumoconioaea g' In addition, occupational exposures have been linked with about 4% to 10% of all cancer deaths,',~ and pollutants with apprvxunately another 2% of all rsncrr deaths! Although evidence is generally uaava.ilsble for the long-term e£fects of ambient pollutants on cardiovascular or pulmonary death rates," aigai$cant el- evations ofrsspirable pollutants such as particulates, eu]fur diaude, and carbon monwdde have been associated srith transient L'lcreases in daily mortality rates of 4% to I646 b°'7 Indoor air may present a greater bur- den of pollutants than outdoor air.N1J Environments.l tobacco srnoke fs an es- tablished carcinogzn," and estianates of radon's contribution to lung cancer deaths range irom about 7000 deaths per year to nearly E4 000 deaths per yearas'a ~~, geophysical fart.o.rs such as backgrovnd ioniz:ing radiation and ul- traviolet light may be accountable for some 3% of cancer dea:ls.' The sum of the boundaries for these estimates approxirnates a range of 57 000 to 108 000 deaths in 1990 related to tnxtc agent exposure. The nonfatal er3ects of to)dc exposures in the environment may present even more widespread conse- quences. For example, fatal lead poi- soning is rsre, but the toll from high blood lead levels may be lifelong learn- ing impairment for some of the more than 23000Q children now exposed to blood lead levels ~tar than 120 ~,.tnoUL (25 ~gldL). Urgent questions aree also raised about environmental cha.Zges such as atmospharic warming and,ozone depletion. Given the nncer- ts,nties related to toxic environmentrl exposures 0-*f1I th2 ubiquitous character of their impact, an even more evmpel- ling challenge than identification of the:ir eurrent mortality burden ia dari$ration of the nature of the issues, the exposure trends, and their likely long-term con- sequences. The $gure of 60000 pre- sented in the Table for estimated total deaths ~rom toxic agents represents the sum of the lower boundaries of various estimates of the contzibution of tazdc agents to deaths frvm cancers and (for occupational exposures only) other dis- e$ses of the lung, cardiovascular, renal, and neurological systems. F1re<arsrss Firaar•ms caused reore than 86 000 deaths aaon,gAmericsns in 1M, includ- ing about 16000 horrdade$, 19000 sui- o;des, and 1400 nnintantionai deaths'" The number of deatha caused by fre- arms Is now higher than those caused by motor vehicle erashes in five states and the District of Cdlumbia (unpublished data, National Cent.er for Health Statis- tfea, Septerabez• 8,1983). Compartsan data indicate that fzesrrn-ralated homidde rates frsyaungmales in the United States are 12 to 273 tiTnes the rates in other fndustrialized nationa, R-hereas nar--f•re- arm-related homicide rates are 1A to 9.2 times gre.ater than tbose e]sewherea61 For example, in 1988 there were 1043 IIrearna- rYlated homicides among US males aged 15 to 19 yeare, compared with six such d eaths in Canada and two in Japan.." Ftre- srm-related deaths now comprise 11% of all rhildhood deatl^a and 17% for those JA~ Nbeernber 10. 199.1-vo1 270. No. 16 ' Actjal Cauasa or Oeat~ vi t,e Unhed Statee--~cGt-nls & Fca!;e 2209

StIVi b, •A erux lelecopier iuci i1- 4-5~ + 1u•48 seed 1fi to 19 years, inclnding 41% of deaths among blaek males of this age.~' Firearm•related suiddes among bla.ck teenage males aged 15 to 19 years doubled from 19Hffi to 1987, and aitbough the rate for white maleds the same age did not changa substantia.lly du rin~this period, it was neariy twice as high The ri.sk of siridd e among adolescents has been found to be ne.arly three tim.es greater in homes where a gun is kEpt.'-' Moreover, guns kept in homes as protection have been fiund ta be several tiznes more likely to 1di1 s fa.may me=nbe:r than an intruder.12 The prominent, detzirnantal et7ect of fire- a:ms on overall death ratzs in the Un9ted States is unique in comparison with other countries.8' Sexual Behavlor Unprotected sexual intercourse was accountable for approximately 30 000 deaths in 1930. Sexual behavior is aseo- dated with substantisli,y incea`aed risk for preve:ntable dis~se and disability and ia the sota=e of some of today's raost prominent sodal chpIlenges. Each year, 12 mtZlion persons (two thirds of whom are under 2b years of age) are newly iafected with a se>; ua11y trrsiamitted dis- ease..u An estimated 56% of aIl pregnan- cies among US women are unintended,21 includ:ng raost of the 1 rmU on that occur among US teenagers each year.°D One of the iaoet rapidly increasir,g causes of ae- rious iilness is hepatitis B infectf on, of which about a third is est3mated to be s€acuall,y transnmtted.ZO Among xromen, pelvic 1T1Lfl****+.atery disease is a severe oomplicat,ion of lower genital tract infec- tions snch as gonorrhea and chlamydia Earh year pelvic inflaznmatory disease af1ect$ aa estaaated 1 millioa US women,T' of whom perhaps as many as 1b0 000 beeam e stsn7 e as a result~l The 80 040 deaths in 1990 attributed in the Table to unprotested sexual in- tercourse inelude apprwdmately 5000 from ezcess infant mortaHty rates among those whose pregnandes evere unin- tended,': 4000 from cervical cancer,"x 1800 fram sexualky acquired hepatitis B infection,*" and 21000 from sexually acquired HN infectfoai0 As indicated by the nearly 20% increase over deaths in the previous year flrom sexually ac- quired HIV infection, unprotected in- tercourse nowrepresents one ofthe mcest rapidly increasing eauses of death in the country. Motor Yehtoies Motor vehicle inji:ries to passengers and pedestr•iana caused about 47 000 deaths in 1998' Neariy 4096 of all deaths asaQng those a.gvd 15 to 24 years were caused by motor vahicles,g The chanees of surviving a serious motrs vehicle cresh are increased aeveralfold if an occupant is protected. Lap and ahoulderbelts have been shown to reduce the risk of deat5 by sbont 45% to 65%, and of serious injury by about 40% to bb4o `0-n Airba,3s have been shown to yield a 30% reduc- tlon in fatalities and a 85% reduction in serious irjury in frontal crashea.4 Child passenger restzaints can reduce fataii• ties by 50% to 90%.`1 Use of motorcycle helmets caa reduce fatalitiea by 80% and serious head injuries by 7b4b.u The es- tamate of 25 000 deaths attributed in the Table to motor vehicles does not indude those already reeo..rded as relating to alcohol or drug nse.'u'-'1."3 IuEcTt Use of Drugs Approcdaately 20 000 deaths were caused in 1990 by tilicit use of dnige. It is estimated that some S million people in the United Stafes have serious dnugprob- lems." Iliidt use of drugs aantnbu,zs to infant deaths and to deaths repozted for $uch causee as overdose, suicide, homi- dde, motar vehiele ir~jury, $IV iafec- tion, pneumor,.ia, hepatitis, and eadors tdi- tis. In 1990, apprwdmateiy 9000 deaths r2lionwide were attributed to Elidt use of drugs (b+oth legal and i7legrl) by vital statistics reports. This figure, however, 'does not include those ind4rectlyrelatrd, such as deaths from ard.dents, homicides, infections with HIY, and hepatitis.l In 1990, approzimately 9000 HIV deaths resulteclfromintzavenous druZuse (20% more than 19$9),'~ as did at lQast another 1800 hepatitis B-related deszha'1,11''° In addition, the Nationai Hi,ghway'Iraffic Eafe'y Administration estimated in 1988 that other drugs, often in association with alcohol, may be a faztor in 1046 to 22% of highu-ay crashea" The problem of acru- rately idendfyi,ng d.rug-related deaths is illustrated by a study of the Drug Abuse Warning Net-ark (DAWN) data aatl the national rital statistics repo.rta of mra5ne- related deaths in 25 metropolitan areas, F-hich found that about 7a%more coer,ine= related deaths wese reported by DAWN than by the vital ststiatics sygtera from 14ffi to 1988.1'~11 A study of deaths from 1878 to 1988 in New Yorx City itlentified 1091 deaths in 1986 as `~araotica-rala#s3," only 247 of which had been 5pezL!1cLll,y attzibuted by vital statistics to dnzgr, Some 241 deaths were attribui.ed to un- speei$ed pneuznonia,1'?2 to liver disease, and 118 to e.ndocarditis.ll The findings of that study suggest that there may be a sutstantial underemunt of the role of in- traveaoue drug abuse rrlative to theae and other causes of death. Although Iocal vital statistica reports indicated an in- crease of 50%in d=ug-ralsted deaths from 1978 to 1986, the study cited found a much more rapid incresae of more than 400% for the same peiiod.° Further com- -4 4y [1J~ ~~~do2+r o plicating this analysis is the fact that the use of Elegal drugs by pregnant women increasea the risk for a poor pregnancy autcome, including infnntdesLh,. The Na- tional Commiss;on to Prevent Infant Mor tality reported in 1992 that such drugs may be used by as many as one in five pregnant women nationwide,' The 20 000 deaths attributed in Lhe Table to drug use repreaeats deaths reported to the vital ataUsstics system as dntg-related, as well as those firom drug-related HN usfection, antomobne infuries, and hepa- titis infectiom It, too, is expected to in- c-aase subst$ntially in future years as a result of HIV deaths associated with in- tl8venoub drug t7b2. Qther Factors Lack of access to a reliable source of primary cate is also associated with an inereased risk of death fmm a variety of e:usse8, although quantifys.ng the impact is complirated by the challenges of ap- propriately charsctarizing the varions el- ementa of access an.d distinguis,hing their e$ects on a given health outcome from other confounding variables. Compari- sons of the health status proSles of vari- ous developed countries s~ that resi- dents of cauntries that provide relatively greater aocsss to a full range of primary care sen'icQS generally fare better than residents of countries with poorer ao- c~ess.0 The Cartprr Center project esti- mated that gaps in p:rimary c$re, as in- dkat.ed by lack of accesa to standard pri- mary care, screening, and preventive in- tersentions, aarounted for 7% of prematL=e deaths std 15% of potential years of life lost before age 65 in 2980, of which substantial portions were due to m&nt dest~s! Limitationa on ar~s and use of appropriate primary csre services require very cloee s=tiny as important determmant8 of he.nuth atatus for rna.rLv Araericans amd preseat an obvious tar- get of oppartimity for a nation with 15% of the population curnently uni-iaurrd." Ptverty too has its own direct effect on mortality rates, although it is difficsult to separate the etfect of lack of AccesA to primary care from that of aocial and eco- normic atatus. In the United Siag3om, which guarsntee$ nniversai aac~ss to ser vicee, a subatalrtlal diff'8rt;nt.Lal ren+A;nQ in health status outcome& by social class despite improved accsss °Ji and overall scozas in health sta.tus indicators are soxaetrliAt IucPer thaa those ior other more socially homogeneons western Eu- ropeaa countriea° Similarly, reparts in- djcate that poor ['gT9d{9TK have a pro- jected 11 fewer years of disabil.ity-fiee life than the's more affiuent counterparte, despite gus,ranteed access to medical =rO Several studies that have oan- tro.llled for other risk faetars have shown ! z210 .l4Ma rbvember 10, 1983--vor 270. No. 19 ~ue c,auses or Des,.~ in t-e urilred &a~;nis & Fcsge

SENT BY:Xerox Telecopier I e 7021 ;12-14-93 ; 10:L9 that pnpulations charactezized by lcw educaf3Cnai or ineome status experience poorer health prospects."x" People who are ponr have higher mortality rates for hesrt disease, diabetes mellitus, high blood pressure, lung cancer, neural tube defects, !n juriea, and low birth weight, as well as lower survival rates from breast cancer and heart attacks ,""~ For ez- ample, a study of the relative contribu- tion of various risk factors and income levels to mortality arr,ong blacks eeti- mated that 8896 of excess mortality could be aoeounted for by family income and 8196 by six risk factors (smoking status, blood pressure, cholesterol level, body masa index, alcohol use, and di•abetes), with 8196 reassining unexplaiaed-' Ef- forts to improve health must take into aecoslnt the spetial challenges to those who are poor. COMMENT AND CONCLl3S1ONS Approxisnately half of 211 deaths that occuned among US residents in 1990 could be attributed to the factors iden- ti{ied. Despite their spproximatP nature, the eatiraates presented here hold im- pliatiam fz~r program prioritias. At the most basic level, they compel exarnina-n tion of the way the United States tracks its health status. Clearly, there is a need to improve the assessment of the con- tn'butory effects of etiologic factors on deaths amrrng US residents and to clarify the role of factors such as poverty and restzieted access to health se.n-ices.• There is also a need to look more spe- cifically at how these factors affect •the 5056 of all deaths that occur before age ?b. Moreover, there is a need to assess how they a.ffect our measures on the increasIngly important dimer~sions of morbidity and quality of life..Our na- tional etFiciency in changing the health profile is dependent on our ability to identify and monitor trends for the ma- jor factors that give direct shape to that pro5le. The most important implications of this asaeasment of the actual causes of death in the United States are found in the way the. nation aIlocates its social resources and shspea its progr= em- phases. In 1893, health care costs in the United States are expected to reach ap_ proadmately $900 billion,* an average of more than $14 000 annuail,v for each fam- ily of four, if equally aLlocatEd arross the population. The preponderance of thie expenditure will be devoted to tz•eat- ment of conditions ultimately recorded on death eertificates as the nation's lead- ing killera. Only a srr.all fi-a•ction will gn the control of mnny of the factorr that the Table indicates imposed a subatan- tial public health burden. The national ij,;vQStrnent in prevention ie esti.*nated JAMk hbremDar 10. 1993-VO+ Z70. No. 18 at less than 5% of the total annual health car•e cost.9D There can be no iilusions about the difficulty of the chaIlenges in changing the impact these factors have on health s`,.atus. Of those identiSed here, the three leading causes of death-tobacco, diet and activity patterns, and alcohoi-4re all rooted in behavioral choices. Behav- ioral change is motivated not by know1- edge alone, but also by a supportive so- cial entidrvnment and the availability of facilitative services. The mast rapidly increasing among these causes of death~e.xuzl behaviorand illicit use of drugs-take place behind closed doara and are di?ffcvlt to eon±rront dilectly even in a putatively open society. Several of the causes of death, such as SrP.arms, are the focal point of powerful lobbies that impede constnuctive exploration of the full range of social options. Nonetheless, the central public heath focus for each of these factors must be the paaaiblZity for iznprovement. C•hanee can occvr. In reoent years, trends have been salutary an several dimensions, eg, reductions in tobacco use, saturated fat consttznption, and motor vehicle fatali- ties. The discouraging trends a ith re- spect to the efl'ects of sexual behavior, firearms, and illicit use of drugs need not be inexorable. If the nation is to achieve its i'ull potential for better health, public policy must focus direttly and actively on those fac+.,ars that represent the root determinants of death and disability. We gra •~'y acL--cwla'ga tha i:ralnibk v~- tztbabon to th. prspa.ration of t3a IIssnucaig svdt by ChristL-a R'yp; ewsid, M4H, aoc.' Daad S.veyoa of tt e G'!nce of Diatare Prevent~an a: d Hedth Ptoraoiior., US Deparacent af 33azILh tad H=sa•a 6a-vdees, Wsshangte-z_ DC. We mould tlao Lie to thaak John C. BaII, r, MD, PhD, Iastrs Brenlo.e, Am, ]dPFi, and Dc.vtltY P. Rlce for t3a~r helpfiil earamtnta en the mtnus---3PL R.fsrsnar L Nstionil Centar tos HRalQh Statir'd Adsaua Rrport re PtiRO211/o+saZitx S6stfat:ed, isso• F~yasz- vUe, 3Sd: US Drgt of Heilth and 8saaaa 6e,~ 1998- M,Srthly VILi Stt.t'edra Repsit, Vol {l, N o. 7. Z• US Dept af H..lth and Fiur,as Sac+icm.R+- duaing thu HeaLtL. Can&.qsawftcsa of Srnckfn;r i5 Y.a* of Praprsaa--A RtPe*'! Qf tk. Surgeo-n G+*s- eraL Waals:ngton, DC: US Dcpt of Health a,d Hu- ana Ssricar,1%Q. US Dept of HnaJth and Htman Sftr~ publiaattaa CDC a&841L L Feegs R'Ii, Aa1er R W. Ina-o3ucL.a .a+d mah- oda. Asn J A tv Xad 1481,~(suppl}~6 4. 2 m7s RW, Fi.d,..:na DL Grees-eaecocs'' uu.a'- ais pr,ks,ors af Frarr~ d..th in tm Ua:a.d Ststet .im ,7 Pm M.d 18$'1;3(euppI):181-18T. 5. Bo henbarS R N.Oa+ P, Mihl J, $usaett W, Ids 8. c+nec• Am J Prw 3/d 19s7~{aupplk 3a-42. i DoU R, P®ta R. 2'h. Cavsat ojCaxa.r. Qtienn- tai.•'ts Ssttimatu tf Auoidabla Ris+Ee of Caacsr in Us tlniLd 5hssau roday. New York, NY: Oztard Un:vers:ty PresS 1981. 7, Dfilio N. Pivr.adap HeaL'h 2k*a,gh PubC1o Pa1iq,. P':.kielphia, P'L F. A. Davie Co Pubhab.rr, 186L & Wyxdar ZL, Gori Cs. ConCibution of t2u ezri- rorunr•st to c5encer ux~ ea: ea .Pidemlologic as• -4 49 2203 303362;"- 6 atz:.e. J 1Jal1 Canur l%st 1Y 5$ R25~2 9• Hi¢gin.on 7, )dtLr C$ Envirormrntal g*t+esia: raiseoneoptiors snd Ua,italdora to ancer eontroL J Nat! Ca•.,cer IvL 147M1291-1ffi9, 10. Cari GB, RieLtes SI. Maaoemnnmica of dla- ati+prsv.arSnn Sa L4e Unitsd Scstas prqvaat~en ot aujor auxs of mortsL'ty uQUtd altar IL4 tabls ass:.7sptiora and ecmcr c~jec~ as Seiencs 19'S; 200:1124-I130• 11. rdhit. CC, Tolrna D D, Fik-ss SG, MaC,.. D, Carziavu..vl+s disexse_ Arn J Prev M«i 1B8ii;3 (s4ppik4,.4` , It US Pr.vtntiPe Sessieee i ask Foree. Gticidt to Ctinse•alPrn+.nLfsr&erv{aca•Ae.iuw,arm.+tt ytks Effedi.e+s~u pfIQ9 I~lers.rnf:cnt, BaltimCtt, b~~ 5~ Ilisrna dc F7{.lldn~ I989. Ll Gnfdmta I. Coolc £F. The d.elLZS in isrhemic hesrt dieeus rt~art:Lty r~t~ aa a:uly~da otthe en a- puzte ve e.t4ects of roadies'. fr:~ entiena ar,d e24zg.a 1n l:lastyle. Arcrt Intsm Med 1S86;loid~ 14. Pttienbdr~er RS, Hyrie RT, LTSiah C, F41ng AL Phy~dnl aedvity, other i:fastyla patt$nt, estd7o- rasa:kr d7sei.e and longev:ty. Acia Med Soand. 1SS& illttvgpl)86-91. 1E. Smith ESO. The ec,notnlt izzpect of prevant~ ahl. isahacie heszt di~at. Ccn bf ai Asaoe J.19T 111~07-b1Z 18. Musaon JE, Toataban H, Satergeld S, et a1 Th1 p~f pra.a~ efIIn Cf rAf63~"afl iIIfG2~eS~ofL N E+~gl J Xe3 1942,8~i16o6-1416 17. Cantera fiar !>iaaau Control and Prav.ntiost C:Yarettr at~bnt.h .rmar'a]ity and yetrs of po- ta;tai L'fe IatL-Ur1a.d States,1990. 31MWR Mcrb Martal Wkty Rep. 1443 4°•&ir 649. 18. Klaiaasn JC, Pierre sB, Kuiuu JH, Lam GH, ScIlr%a= RrF, The e`iectn of mstaral amck- t-Z an feW and infu:t tno.jslity. Am J Epidem•inL 1958;7.272r4-s32 19. 13eianan JC, A'sdsnr JFi. Tb. a!lata of ma- tarsal r,nal3ne, Physieal etLt..•-•e, and eduestonal ..,r,,,.,ent an the inddarcs of iasw larth.veiQht Am J Epsdemioi 1985 3Y1BS&Sbd. 20. Tba Nati=sa1 ~~~**!~ or to Pt•eoett Inisot Idot.allq. I`raublirep 7"nd•r Persiat Sho+tehenp• iny Amsri=a Nac G.romtim Waahiagton, DC: US Escitor.aentsl Prvt.cidoa tjenq; zhe Nt- L+onal Camaisaioc to Pr.v.,n In&Lnt Mert.vity, )d.arch 19 ^' 2L US Envitonzosnttl Fsvts~on Ajiiar~. Riopt- ra.vry Xscttth Ef.Ga of Paariot Pr.okinfl, Luz4 Can.-rr and Ot3+cr D ~a~rdsr~ Washiay Can DC 199Y- U8 £nr'avnmantsl Prvtes.tien Ag~ publimtion 604$ 90A63F, 2Z Shu}tc JDS, Navotny TE, Riy DP, Qn•nntit)in3 the d:asaa. imP.et cl~t'.a ~n}~.ritb E: Jti+- 3SEC II saR•azs. Public FLcL'k R+p. 1991;1OH: 32b-SM Y8 Hatotu: Ctau: for Hecw,h S•K$t1M Aaalta, US, 1sai FT,pass•sv£le, Md: US Depc of Fialth and Hnrrsn Sarv'.rss; 1983 US Dept of FTeaYth and Ho- nsa $er-~eu publ3ozion PHS 93-1M 2l. US D.pt of 8eslth and fIL-aaa Errviem ?'he Ssvyeorz Grwral5 R•rro•rt oa h~b-;rian cuS Heullk Wsahingtan, DC: US Dept ot Health and Hi:assn 6sniam 1988. US Dept otHtei:h a.nd Ht man fier- v;ces pnbt{mtion PHS 88-5=lA !5. Pa3snbarYer Rd, ldtivde RT, R'usg AL Physiol sedvtty u,3 physies3 IIaes s dstnrxnLunts of b.al th a:m+..' longeFiz~r. IsL 8oacu,ard C, Sb.pl srd it7, 3tephens T, 8ur..on J'R MePDezaon BD, edAL Ea- erd.., t'it+usr, aud fl.altl 0am,-aigtt 11 Humo.z R7net•ias Books; 199Q• 4E. Po.til RE, Czr.. wa CJ, Hoplvn JP, •Ford ES. Y}u'Z1ca: ecl-1vi.,y and e}.xanic dit+as&s Am J CciR Nr1s. 1~107i 27. PoseD liE, ThampeC= PD, Caspersen CJ, Hen- d-^kk J8 Pkyaic+l aedvf•.y .ad the tacidenet of ooro- aary hesrt diseat,a. Ana Rro Ptibtie fleult3. 1987; a158-s87. Z& Lee I, Psff.n3uyQ R8, Hsieh C. Nsial aa L+vity a.~d rieic of davelap's~ mlora~.a1 n~e.r anone ooi2~. alszatsL J Nati Cax.ear In.et 19810:13YR- t9. H.ha RA, Tcuuch 6M, Acr.h.- bsrg RB, bfarib J3. Ls<eaa deszha from nLnt eitrorlc dla&saee in tha Acnla! Caisses ol Death in rie llnr.eC Stl~v-~.s-^;rrxs & Foege 2211

Iv12GUP1e( lULI +IL 14 `JJ + IU'.1 Q 4 4 v~ ~V 1~ %? vc++r I United Statca, 1986. JAMA 1~4_Bb6-E4d3. 3d StartlaJ, Rcer G, S•amla R, ElluRt P, Dyer • A, A{ru-tsod M. LNTERSALT stsdy 5nding& publie be:tth ard medicsl nre fapliea;3o;... Hypntrn- riwc 1A96;14374577. 3L Lav MR, i3'zoet CD, Wald NJ. By bua mueh doaw dietiry m+lt l.duction Iawar blood prer,aure7 III: =aal,yeta ot da:a i:orn tx'sit of ta7t r.durSon. B1dJ. 1991r6t-_,411-824. S2 Fierrau WH, Teutstb SM, Gaiaara LS Diabe- tan meIHSSS Am I Aeu Af.d 12478(suPD1):7P-8` it Hendasaon BE, Row RA, Pika HC, Tcwud the primary p.•s+sntinn of anetr, 5es~ 1991; ?5d:I1d1-1188. 34. DeIl R. Tb. Lesacna af lf.ie: kayrxr-e addtas to Lht nutritioa and c°^^a* QonftrttlC9. CaRCfr R1J, 1S~attypll~3. ES Newb.rn. Pbi Diit and nutrition. Bull N Y Aead Jd.d 1975,3(S8S3.r+8 Sb M.ason JE, Na..tua DM, Ktol.a-slu AS, Sta~ar MJ, RTMett WC, 8enn3uns CH. A pre- s-•eedw atudy of etardaa and inddenae of diabetea amosLg US male plLps:eiuw. JAMA.1@:I'2•,2M-6&67. 87. US Dept of Eisalth and 8um:a Sc-vicea. Srv nti;h Sp.ia! R.port so t is US Canprw rm AIeo1~o! andHc+sllh Wes}sieYtaty DC: Nstioail tas iNte an Aleohol Abuae u+d Aleobollsrr~ 18f4. US Dept oi Health and Humus S.rvleu gublimL'on ADM So- 16FiB. ES, Sebx~ CA. Expaa9rs to slcohoL'm 1n the fsrm'}y--Unit.d oJVt es 1888. ALVawre Daja hvrx Vital and Hmltk S 'a No. dt3S, L vat:an7le, Md: Natlanal Cesta for $ealth Statsatlos;1991. U S Dept of Health and Hanaa Sernie+a Pubtia,tion PHS 91-19SO. S9. Waat LJ, Mal.well DS,?doble FP, Solomon DH. Alcoholisa. Anx lxterx Msd. 18S4;1oot4~18. dG. Johczne. RS, SshanQ S.'`', 3ieadein? AI, gu% su J, Roth HF. Digestlve dis.asts Arn I Prev Md 1437~(eupplkS9 SFi 41- MeCay GF, Jo}t'utaze Fth. N.Loa 1W, DutfSe R,B, A Pe.}em of ltts1 toad sedd.nts in O~dardstSrs aa.r a &yer =esiod.lr,{-+y. I~b-68. at Smtth GS, Fs.ih EL Uaintear'.c4a1 in7nr4as. Am .7 Prsv M.d 1487 3(axpgUJ aJ_163. a Csntara !ar Disaaae ContzcL AlcahoI-rrli:.d moztaltty and ymin of patantial tife lost-Unitrd Ststsa,1987. MddWk Mvrb ,1do'rial Wk1y Rap, I9'40; 88.178-175 43. B®.nett JV, Hohra~erY SD, Ropexs 10', S¢ Iemon SL. Ir.f.e:iona and paras•tic diseasaa Am I Pnu b1.3 1987~(s:ppi IU~lls tS Nati onal Cantel for Hsslth S tatlsd ez Vital Sia- rfativpfLlt Unffed SlaL+-1 F90, Volumt 11.• bCor tzliGr. $pamsalle, 2Sd: US Dept of I:Io.ltb and Hurnaa Sarvics. In p:est 46, R''ilisms W W, Hic]aan 3i.A, B- ne MA, Krndsl AP, Sp4ca JS, H' n AR. Ia.atW utian palSdae aad r.ccu* covrragr amos~ achilts Arns I+cra-x M+d 198$139:615-625. 47. C.atera fotDisLL+e CoatroLACIP.reaornrneo- ds dots for protsedoa sgataat viral h.patitia ki3CWR 3lorb Mortal Wkly Rtp. 18&SS4B182~i 4d. D; BiscaQ'3. Alis. C7uoaie s'sai hapatitia and bs,patnoe,IIulsr esr+s+os.a L^, the Untted Ststes. In: TaLcr £, IX Btwqe;- 1X Psuaan RH, .ds 8'ti- elo9Y, Patholo,y, and T~sazsrfert! ofHtpatocellular Careirona fn North ArvrCea. Houyton. Tez Part- folfa Pvblia6ing Co 1989. 48. Jereb JA. Kally GD, Denley SW, Caatbea GM, Snldsr DE. Tt:bsroiloaic tnorbidlty in th. Urssted Stata& 5na1 datt. 1992 MMWR Mo+3 AfarmJ Zi't-ty P.p 1991 4PY~27. 54 Csrrtas for Dia..ee Ccsa-rol arsd Pmvvsdm. HN/ AIDS Svrvr!llaxn R~ Atlanta, GL US Dept of Healtb and Hzaun Serr~ Febrs:ary 1438 bL Parsa FP, Rs9 DP, Ws**wte{n IB. Caror.o- psneaia m.eLv i=s: the del+ate cent:auae. Sci.raA 1991}S+'~03-9~b1 bL Fater LS, SchieBdbsa CW. Resplra nry die- aasas. Arx J Prea M.d 18B7Xanpp 1,kS1S•12f. SL i&adzipan PL, Mar3m+its S. Gurrsat magni- rade of ocouparlan.l dSa9sse tn the Un3ted Rsuea. l.nn N 1',4ead Sct. 1983Ji7337•1b. SS. Hammond EC, GarQnkal L General air poihr tion aad anear In the United Ststa Pres, ldob I~~621L 64, Pape CA, Sck ntz J, Rarisom MR Duyy mar- tslity snd PH1Q poilvtict ie Utah vaIl.y. ArrJc Enviren H.aJalt 1419;47511-217. bB. SehWastt J, Dodcezy D. PartSculat. ai.r peltu- tion and da'.1y tnortality in 8teubeavil* Ohin. Am .7 Epidemlol 1M185:12-ii1, S7. Wichmsny HE, kneller W, Alihoff P, et a1. Hoaltl: efCects durL4g a s` c+g epiaaie !n Wert Ger- taaay fn 198d. b'vsa;'lo+t Health Peraprri 1f&?;72 89-99. 6S Swansan GV C1ncer pr.vsntion in the .ork- placY and tstural envisonmKit a review of .tlo1- ozy, ressreh de2m and meihada of siak rsduc tion. Ca?wr. 19~ 62172:.174~ 59. Willate LA. EPA Pro}SCt JSutnrsary: Trls To- ta.I Rsposxr® Asuas+n.est Md.hcdatoyy (TEAM) Study, WashinQtna, DC: US Envlromaasi P:oteo. tion AgenSy, 188a 6+1 Pualan J8, Nalacn C8. EPA'i pt-spec t ve on riaks from resideat:al radon esTxm:ra JAir Wasts Maacgs As.z 196_°A9~91Sfl21. 61. Lnbin TH. Baiee JD. Er'' g Rn-induesd ?ung cancrs in L4e Un lie3 Ssslaa H.eLtk PAys 1989y57;417127, 62. Agenq tor To:dt Sube<snees tad Diaeaae Rag- ittr7. Naturet and Sz'sxt of Itad Pot.o+riag i,% ChiZdre+e {n the US• Repv+t to Caw4rns Bethes°s, Aid: US Dept of Heslth asd Husata SCrL-eq JuIJ* 1488. US Dept of Iiks'.th ind Hraua Strvic,3e pub• tkatjon DIiFiB 9S,?1.'93. 6& R'ebatsr D W, Chulk CF. Te.ret SP, R'lr ta mvte C+T. Redtsd-,g firei.-^ ir~ariw. laasw Sof t.rJj- fwL 1991:TfFr79. 61. Fingarhut LA, K1•'tc:an JC. Ftreua mmta]y fty a.-aoaY ehiZdmrn nnd youth. Ad.cux Data Frons Vi:aI and Hsalrh Stasitei;t No. 17B. F%~ Md: National Center far fieelth Statistis 1959, US Dept of Haalth and Hr=sn Ssrnicea publintion PHS 90-1250. 66. Sloan JH, Rfva.z FP, Rea,y DT, Ftrria J1J, Pau MRC> K.11a---a-n AL. Fareirsf r.Sulariona ard rarea of euScde: a oompazSLOn of taro met-o- potitan araaa N Er,q2 J M.d. 159PS2~-373. bS B:mt DA, PerpAJA, Gol~ CE, ti aL Rtak i`actass for adolesmat cuiddd a wipu-Iaon of ado- Ieseent tuidde tiiat;ms +r?th suiddal!npaaenta ArtA Ge+t Payrhia-•ry. 16S;96•1+81-588. 67. Fia~erhut L1~ ISlsiraznJC LZtezsat3w^sa] and ;Ttarttate evmpaziaora e 3omidde naoc,Y yecng z•salen_ IAMA. 1~f~?`1S19210,2Z11. 8Q Div3~ ofSi NiTlv PrsvsnCan_ Sanatly 7>+ar mi~ Disncar S~.crcaillarxc,lP9o. Atlanty, Ga: C.n- tea Lx Dieess. Ccm-_at US Dept of Hes1h a:d Humaa Sexvio=a., Psblia HesSth Servkr, July 1991. 60. A.at.tiaw SK Van roet J. T.onagm abartia.n, birth and gregainp' str.da^iR an stpdate. Tom Piante Prrrpsrt 1889•~1M.S& 74. Altaz MJ, H.dlar SC, lfirsaris HS, et sL T}:• changing epidaniology of hepatttia B in the Un1tLd 8tata&- need foz alurnative vaedaaacn atrateg~as JA3lA. 199p',L~I21S SY?L 7I. Waahiff8',.oa AS, Amn PS, Broala ifA. The •moie eosu of pelvie lz~; dir.aa-e JAMA 19E~55178b 1738 72 Centers for Diaesae CentxoL Divisio+e QfSM HIV Prrsvntforn 19G1: A+snual Report Atlanta, Gc US Dept of Haa7th aad Hurws: 6arrim% iBM 7E. Wrsid Iiuahh Organizstan CoDabnn7ag Gn- Ler In Pgrssatsl Cars avid Heal:h E+.^Kr¢ Rd iitth in X•+• *mal and Chlid Cara Uni srad.d preQas.-p' and Ir.fxnt mor•.al3tyhaarL-Sdity. Asn I Pr.v Af.d 1se7;3(ssppll130-l4? 7d. Am.rian Canrar Soc.:y. Ca.err Faeta and Fig~1D9J. Atl:nta. Ga. Amerisan Car,cnr So- eiety; 199s. 7E. DeVi's YT, He1aa3 S, Aoaerlrly SA. Ca+scrr. PrircipLu and Pwliee of Onmlcgy, vol 1. 3td .d Plu'ladelph3a, Pi JB LipF,...ro¢ Ca; 1M 76. N.leon JIi, Av.zeCe IiE, Ziehart RM. Ce3vi- esl intr..pitheHal n.nplama (dyaplasia aad csrd- noaa in r3LU) ard e.rky ia,rssre oerrk-al ca*ciz,•••., CA Cauar I Clvs. 1589;2:k1S7-178, 77. Campbell B,J. Satrty beft h+fary aductinn r®- Istrd to awh severity .nd front ssatPd p¢sitSan I 1 rau+na. 1957M7SS-74. 7E, US Dept of 1`rutspnrtation. Ua of Co$troll.d s Su6si¢nsse oed Xipkuay Safe{y-A Refzrrrt la Cui. gnsac W.atll-rn, DC: Nat!oral Righ®ay Trazr Safetp Aximinlatraton; btsrah 1II68. 79. A1mhoL Drug Abume, and ldeatnl $ea2th bd- m's.:atratitm_ Drup Abuae and Z?ny Abust Reeesrek. TJu Tkird T4i.n~t{al R~M to Coxprau FYe»rs tti. Sacrrfary, 73cpartmnst qt XcrsLUt and f3u*+aa Ser. afess. Waahi~ DC US Dept of Heatth sod Hus,an Sarnica; 1991. 84 NaiSonal Ina..•'Cte on DsvgALca&& AnnualLzla, 19&9: LSafa f'Rrrn Lfce Drrp Ab%u,s WcrniaQ Ne4- smric (DAWN) Si.i.r 1, Na P. Wsth,iagt,n, DCc US Dept of Health and Hu.man Services;19A0. US Dept of Health and Huusu 5ev{crs publimt;on ADId 94-1727. Si. Pollsrk DA, Flol:xs8-een P. Lui I{, Eirk hiL D~€pt.r des in the repcYed 4equency of wsinc rsla.ted desta, Un4ted S'1us, 1958 througb ISM JA1NA_ i9BIA98:2M8-2W7. S2 StanebusgsRhlSesJL-°s1sDC,Be.-eaiD,et aL A lsrysr apacetrnm of s.v.re HiZ7-l-rdated d3s- eaee in intravtnous dnig Wn in New York City, Sessn es. 188~' ~ 16-919. 91 Snsl.ld B. Primssy osre and heahh; a cros.. aaaonsl eompariaar IAMrL 1991,'i66~£$ 22I1. 8s. EmployeaBanefftaRea.arciilnstitut.CEBRI} Ax EBRI Speeial Repors and lasw Brlqf Number 138. Wsshington, DC: Empl¢ysa BenaEta Reaessch Institsta; Fehroary 1?9Y.. 85. Mu'imt 2r:G, Smita GD, SwsaFald 8, et al Halis .qtu8ties amanY Britiah dv1 ursaats ttu WLitt t;=~j. LcusL 1fi9S;$$'!:12ki-1399. 88. Terri: bi Ltasons frotn Caaads'a beahh pxo- Qnm J FscbL{n Heal:k Palicy. 199021E1-160. E7. Tyzoler ldA So~oecnr.amio atatz:a in th. daasiology and ttvaiinent of hvp.ztaesaiap Hypsr- t.n.rm 195tP13(s~ppl IkI94-I97. 88. Salonen JT. Sodoecona:a;e atat7ia and riak of eanrer, eezebrsl stroks, and daath due to enroruy baart diaata* and any dnr,saa: a lcnsitudinal s:udy !n es~ern F'uland. JEpi3snsld Commu+dty Ractt3i 1~3629G-291. 89. Daral HH, Po-es R*i, Cniu C. Rsoe and ts eioecc,amle Ststua in r,:rsical -'rom brBa.st es.tior. J Chroxie Dfb 19=M{7-.-6'3. 90. Haaa M, Kapka GA, Cat:utbD T. PovtT-y and heal:h: pzsspetlre evides.ee `.rom the Al¢m.dde Coct:ty Study. Am J Ep i1s+r+foC 1957;125.-9SS-993, DL Baquet CR, Horm ?W, Gfbhs T, Greaiaald P. 6x~o.eo:iom3r taetera and ei.x incdena among b:atYs and Whitea. J Natl Canttr IszG 1991 ffi So 1- 557, 92 x'lai Llry TdAJatilia DE, Frank E, Fbrtna.nn SP. Scdoeconorcie va.^na and haalth: how .duts- tion, 3acoma1 and oecup.aaa mntrf? a:.e to risk faa toca far ordaavaan Lr di~s~ Am I Put+lic H.al:i. 1944~LSI6$Y0, 43, Gu.-alia.k JM, Land KC, Blsaer D, FQlanl~um GG, Branr3h LG. Ednca: ans7 at--s and actiPe L'fe axp.ctr.ncy uarnsg older blaeks and vhit.a. N_r+spd I3[sd. 198s SE911a11B 94, PBppa. G, Qceea S, Hadden W, Fiaber G. Ttu fh•c'sainj dlsFarity in mortaL't;y betwsn aodoeca- nomia groups In the Unfted Sta~, 1660 and 19Bfi. N S+y1 .7 M.d, 1@98;8"A.1QS-I08. 4i Y•aplsn GA, Fiaaa MN, Syme SL, M3nYlar M, W]:L~b' M. $oda.cvnamie sutaa and hatlth Am J Pnv.rd Aled Il~S7;atauPp1ki1~1~. 9d US Dept of Health aad Hnasa Scvioaa. Re- part of L1u S.rnRury a 7 atk Foree o~a BJae1c and 3fiivxtry H.alth WoshinCtotL D{ US Dtpt of Health a.nd Hus,ta Svvioes; 1988. 97. Cr^..ea 9f W, TeutaeL Sl{ ~YQLursaoa DF, D:ari® JS The eZ.et of laovrn riak bcy.drs on tke exesaa a+ortslity of blark .dulu ia th. United Stas..b JAdIA. 1D90~3$~5 8 D 9B B:aaar ST, Waldo DR, McSnsiek DR. Nuda.-ul hoahh axPanditures prqjatiana t}lougb 2030. Hsaltk Cars flnau Rn. 19f.?14:1-I6 93. C-tas S6f' Diaeam Cantrol Es~, tsa ea natianal spsndiny on pr.vnnt3ntt-Uni:zd Statea, 19E& M.YWR Mosb Alo+tia! W'kiy Rqs. 199p~/1c5LE 6SL t 2212 JAMO. 1sa"nber 10, 1993--VN 270. No. 18 A(;h;aJ Ca~ssa a+ Deatl ;n 7)e llnhaC SMIas-Mcalrvits & Foege