Tobacco Documents Online

578 In a much larger study, mv colleagues and I_:amined the relationshio of environmental and ethnic factors to cancer mortality by site for every state and region of the U. S. for a twenty year period. We began with a very large number of vari- abies or factors, from which we selected 12, among them c'-garette consumption, intake of different kinds of alcoholic beverages, ethnic background, oollution, income, temperature, etc. We used these to derive sophisticated statistical models of the geo- graphic patterns of cancer mortality in the U. S. These models were an attempt to describe b_v statistical ecuaticns the relationships between certain factors and cancer mortality. Gie found that all the factcrs used in the study accounted for only 36 percent of the geographic variation ir, white female death rates from cancer of the bronchus and lunc speci;ied as primary. °he best model -- which eliminated the unimportant factors and brought into focus the important ones -- accounted for 29 percent. (In contrast, the models accounted for 89.5 percent of the variation in cancer of the breast.) This indicates that unknown but crucial =actors were not included in the variable cool from which the model =or primary lung cancer was derived. Cigaret*_e consumption, which was included, turned out not to be a factor. We also disc inarv lung cancer _- -- 'or both white fer ..ecause it shows we cE ?resent a factual pict There are ot .a'_anced discussion of _. '_ncreasing numbers __.__- have traditional: °-^aerlv reserved for =_-_^un'_tv and work en _.. -.-„wn carcinogenic c, The tremendc _rtorne chemicals orcc ou tdoor anvironme:: -';esc a set of facto ay. In 1976, I _ ~cssib'_e influence .c•,:ston upon lung c_- <--_atorv diseases. facili=ies, mort: _:.seases were fcL -aa_is =rom respi=

579 colleagues and I examined d ethnic _°act.,rs to cancer 1 region of the U. S. =or a •:ery large number of vari- ;ted 12, amonc them cigarette .ds of alcoholic beverages, , temperature, etc. Ne used 3tical models of the gec- in the U. S. These models atistical ecuaticns the 3 and cancer mortality. aors used in the studv .e gecgraphic variation in - of the bronchus and lunc el -- which eliminated the focus the important ones -- ast, the mode_s accounted for ncer of the breast. ) This 'actors were not included in iei for primary lung cancer , which was included, turned We also discovered a•aarn'_ng in the basic data dividing __imarv lung cancer from the unspeci'ied category of 1•sne cancer -- for both white females and white males. This is imoortant because it shows we can't be sure lung cancer rates, as reported, resent a factual oicture. There are other important considerations _ cuired in a balanced discussion of lung cancer in women. For instance, women iz increasing numbers have, over the last 35 years, :eft the jobs _:ey have traditionally performed and have entered wcr% areas 'orierlv reserved for men. Wome.^n are now esxsed to _=:e same commun'_ty and work environments as men, and there are many .n:cncwn carcinogenic potentials i.^n these env'-ronments. The tremendous varietv of toxic and carcinccenic airborne chemicals preduced by industry and released into _.^.docr and outdoor environments, in our homes and our '_aces, suggest a set of factors which may be .^tertant _.. '.ur.g cancer etiology. In 1976, 1 published a 30 year study which examined :e oossib'_e _z'luence of the oetro-chemica'_ _ndustr•• e:aosures in :custon upon lung cancer mortality, heart diseases, and ct:er _esniratorv diseases. In certain areas clustered around _ndus- __ial ?acilities, mortai?ty rates °rom respiratory cancer and .,taer diseases were found to be ver high -- so ..igh, _.. fact, =`ac dea=hs °rcr: _=scirator•i diseases near'_;• doubled over the

580 last 15 year period. Outside the industrial "pathway," morta!ity rates had remained stable for twenty years. A similar _indina was made by Clemmesen in an industrial city in Denmark. These epidemioloa_ical findings are strongly suggestive of the fact that industrial exposure factors underlie the etiology of lung cancer. I might add that I saw no evidence that cigarette smoking could explain the peculiar disease patterns I found in Houston. Our increasing ability to identify industrial carcino- gens has opened many avenues of study and emphasized the _.^..portance of industrial environments. Control of these carcino- gens at their source offers the most positive potential for prevention of lung cancer. To date, epidemiological studies of lung cancer have not generally included adequate in'or:nation about occupationai exposures. Frequently, an epidemiologist may get little more than a brief indication of occupation -- such as, "construction worker." This tells us nothing about that person's actual exposure. At the M. D. lndersor., cne of the major centers in the U. S. for the study and treatment of cancer, records are ever. less informative: the vast majority of cancer patients' reccrds simply list "retired" as the occupation. How helpful, then, can we expect the reccrds of private practitioners and community hosoitals to be? Even when attempts are made to elicit fu'_1 __.::rational expcsure his -.emory of previous expos _^conplete. I seriously .as an exact record of tY :-.ounts, to which he may i :ariod. And his wife, wh •r_usewife, would undoubtec -.._ cleaned from his work Still other imp _.._ aentioned, are emer ,..c__ causation. Scient roles played by _ __ncer. Although this :_':usions do not aooear ._-~a ..= research does poi ___...,c unknowns in t^=_ ,rom mv own stuc _esearch recorted :- _=nt that calling c'_ factor in femal r comolex eoidemic

581 stria: "pathway," :^ortality years. A simi 1ar _indir.g 1 citv in Denmark. These suggestive of the fact that he etiology of lung cancer. at cigarette smoking cou'_d - found in Houston. :entifv industrial carcino- .dy and emphasized the Control of these carcino- positive potential for .idies of lung cancer have :nation about occupational gist may get little more -- such as, "construction it that person's actual of the major centers in the cancer, records are even ~f cancer oatients' records :)n. How helpful, then, can :ctitioners and communitv are made to elicit full occupationai expcsure history, the patient's ancwledge and/cr memory of previous exposure will undoubtedly be faulty and incomplete. I seriously doubt if the typical industrial worker has an exact record of the hazardous substances, much less the amounts, to which he may have been exposed over a 30-40 year time ?eriod. And his wife, who mav well describe her occupation as housewife, would undoubtedly know even less about the substances she cleaned from his work clothes all those years. Still other important areas of research, besides the above mentioned, are emerging in our study of the puzzle of cancer causation. Scientists are beginning to recognize the pcssibie roles played by nutritional imbalances in the et•:oloay of cancer. ?'_thcuQh this ..iscioiine is still in its infanc•i and conclusions do not appear to be immediately forthcoming, this new area of research does point to _he fact that there exist many scientific unknowns in the study of cancer causation. From mv own studies and from a considerable amount of .,:her researca , morted in the "_*_e-ature, _,t is _:^.ere=cre pparent that calling cigarette smoking the sincle na;or causative factor in female lung cancer is an oversimplification .,f a ver_v complex epidemiological problem.

582 Ccnc'_usicn There is a concept widely accegted by publiz: :eal=: schools that if many studies, in themselves not based or. definable nooulations, all arrive at the same _onclusion, _:en that conclusion must be valid. This concept disregards the 'act that the same intrinsic error in method might produce the same result. ~a the first and subsecuent Surceon General 's repcrts, one of the main sets of supDortlve data was based on tC:e surVeV of L.'le American Cancer JecietV. The lnadecuac7 of t.._s sL•=veV recoc.^._Zed and cr~t_ca.iv anal'7Zed bV Tla nv o~ t..,. ~ore^cst was szatistic?ans and epiden'_o:ccis_s at ..i:e time, including 3ir Ronald cisher, the 'a_:er of statistical methods as we ncw _ ractice *_hen, Dr. Joe Ber!cson of the Kavo C'_i n_ ,who devisec the .^..ethcd o_- reDOrtlng su_^:ivals we ai_ use, ?rofessCr K. A. $rcw.^.'_eer for the ?Iaer_can J'tatls.._CCai Association, Dr. DO nald Mainland, ph.sician and teacher of medical statistics, a...: numbers of ct: ers. In the Cancer Society study, thousands of women y cuesticns' about rolunteers asked of their acquaintances man, tr eir tersor.a'_ habits. 'Io one cuestions ...`.e si::ceritv o` _.._ women, but no business would consider ____nc its sales golicy on (: )

582 __.:sion widely accepted bv publi_- , in themselves nc*_ based on .ive at the same conclusion, _ :en This concept disregards t :e 'act iz method might produce the same :secuent Surgeon General's reports, -tive data was based on the survey The inadecuacv of this survey analyzed by many of the `orencst sts at the ti,^e, including of statistical methods as we nCw i of the Mayo C_'inic, who devised ~-als we ali use, 9rofessor R. A. _,.i st'_cai Associa-lon, Dr. Do :a'_d :er ..'_ medical st3_istics, and aty study, thousands of women ;uaintances many cuesticns' about _ cuestions the sincerity of the _onsider basing its sales policv on 583 _..,. results of such a studv. The oroncrt'_on of smokers _..__rviewed bere no resemblance to _he procor_'_on of s...c:cers _.. -..e pepulation, which demcnstrated that the women surveyors Hen_ cut to _ind smokers. Twenty-five states were not :n the study. __ was largely an urban New York, New Jersev pcpulation, c' pcer ...:ddle class people. An enormous literature has grown up on the sub:ect of =...o{inc. But the bulk of the evidence is statistical. _._ nc s_-c:e :nstance has cancer o~ the lung been produced an•:._ean:ng='ll animal experiment us:ng tobacco smoke _._ha_'at'_o n, even _-.,.:cr millions of dollars have oeen sz;ent _n the _==or_. :n a_1 .^.IV Vears as a cancer eCl':emlolCC_st, Z^aSe seen ..._nv tn:eories studied and debated. But Cnlv in the instance of =ne smoking t:eorv has the theorv itself become sacrosanct. A:.•; __ _i:e many vad id scientific observations which _..:..t_ =he t hecr^ __ _..iscounted as bizarre. .%!an y a doctorate on ..uman psycnoloc•: be earned in the future, when the actual cause of '_u::c __._.,ar heccmes known, wnic.- will attemDt to _..,. _..t..nsit_ ' _..e ?romoticn of this unproved -:zecry. Science, by explaining one fracr..ent after another c: .._tic, .,iochemical, somatic, behavicra:, env'_ronmental, and _ nal ~..'crTaticn, ~s building gr-adual:y the structure, (1s) 1 (iti) I

584 cahich, upon co-nleticn, will help us understand __`.e cenesis of cancer. i_ is beceming increasingiy clear that =:e_e is ne si.^g1e, sisplistic answer to the question of what causes respiratory cancer. It is hardly in the best interest of either science cr gcvernment to create the illusion that an attack en a single lifestyle factor will provide the so ution for such a complex problem. The proposed legislation creates this i'lusion and should not be cassed into law. 1. cmba: ^un1 _ :_•~1 '2 ';r 2. J. uG?-i_ . _: ...~ . .. ~C.• . ..a,. :cr. .ssarua?tts. = ~~- - ~-~? -.. 3:Ii1 SGGI -- -' '_..

understand the genesis of clear that there is no iestion of what causes the best interest o: either. .1lusion that an attack on a _ the solution for such a- _ation creates this illusion 585 PUBL.C\T :O :'.; 1. i.cmbard, H. L. and 'lacdr)r•:(•.!, E. .i.. ?'ublic 'r'^alta. Courier o: thc L~~rrnatior.al.~ F•__'ern .,.. .,C ..! :-„•.77,77 -(') ).l'-14, 1131. 2. LomL':irci, El. L. , and !'lacclOnald, C. J.. St -... . Cancer C:inics .is Seen by the Practicing ?hysician. =v:'_zn! " .ocl, "1Oi:(20):9,9-951. 1911. - 3. :.omcard, S. L., and Macdonaid, E. J. : Co~n1et_ res ,:is Control of Cancer. Public Health rsin 1. tac:ionalci, E. J.: Iiistory of the Czncer ..}_sac!•,;:setts. :n Cancer arlc! Other Chronic Discss.s fiuuqi:tcn >tlf:.in Ccicnany , Rivers'-i=. 1')5,). E. .,.. Hi::toricnl Sketches or. for the Pt~ro~r!:~i :Yts D~,o.,rtment o.'~; . i: i; ,•z:cr'.;•-six broa tcast} which was repriater3 in t`e . '"ha :'is•a o` PubLi.. ;:,.a1t): consciou_:._.;s- 1934. E. J.: f(isterical Sketc!:cs c.n . rFpa;.,d for the :•tassachusetts De',ar•mr2^t ! .. ...... ... . .+r.): .c! I,, 11.,. .. ' i.t `i)rl.'-3ia brUarlcasti; which _sn.i J !ei. Edward Jenner, 210:(23):1233-121~9I, ..., - ;. 'ac__n.3i; E. J.. ;!isto• -~3l 5k_,ches n?'_ _- <~1 for the Massachusetts De;,artaent o:.:•~ -_ :!C31ta aad sccnsored by the Mlssac!:usetts Ltedical Soci:t•.•. .r.e of `ort;•-six broadcasts which renrinted in ~nqlanc 'e:. Vaccination of Toda;., 210:(26):13O, 193,. ?. t•!ac:'.~sa1d E. 3.. The !(istorical Trend of iabe-es. Jiabet•_s Vumber, 21: (2) :57-64, 1?.,. 9. ?~~r.onal t E. .7.. f!ist~r:.~l Tror+..lr, in Can..,,.. _c'^ont: _tlth .,=,ncer uu-iber, 21• (~) •.'•37-2G5, 1934. - '_0. :dac.ionaLJ, ,,. J. Chronic i2h„ur.t.ati:•m. Th.= Cc,-r. t_c: ~.dult Hygiene Number, 21:(1):25--'7, 1934. 11. _,:onald, E. J.. Cdoston of 1R00. Tho Ccn. ,a^ .ii ;c.tdio :utab:er, -_. (a) , 1y35. E. J.: Fundamcnt-ils oE Epide:r.io:e-;y. 3adc1:= :r` ~1 19-22, 1136.

586 publications (con' t) -2- Eleanor J. :?ac::o::_-::: 13. Macdonald, E. J.: A History of the Massac-:usct_s Department of Public Health. The Commonhealth, 2: (2;, +3 pages, 1936. 14. Macdonald, E. J.: A tlistoric.ll Trend cf (revia=d). The Commonhealth, Diabetes Number, 24(2), 19:7. Publications (con 27. '•facdona ,Jno !tave :f:id Prov 82-84, 1944. 58. Macdona. Entity for Contro: 1sa- Macdonald, E. J.: Accuracy of the Canc3r Dea!l": Records. P.n J Public Health, 28:(7):918-824, 193d. 29. Kacdona: .iho Have Had Prov h f Health Bulletin, ` e _.~__.. t 16. Macdonald, E. J.: New Activities o of Adu1 Hyr.iene. The Commonheal•`_h, Cancer Number, 30. +lacdonal 25:(1):d-7, 1938. - Needed. Conn Hea1 17. )tacdonald, E. J.: History of the Massaciius~tcs ' 31. Macdonal Cancer ?rogram. -5:(3): The Commonheiith, Cancer Number, for Cancer. Conn 192-202, 1938. 18. :•1acdonald, E. J.: The Evolution of Cancer Contro= in Massachusetts. The Med Woman's J, 45:264-270, 1938. 32. Griswold Connecticut Cancer '`.at Med J ll: (5 19. Macdonald, i. J. : The !iu-3nitarian' Move:cer.t. 33. Macdonal The Comaonheal`h, Radio ;1umS,•er, 26:(3):176-179, 1939. Breast Among Women 20. Macdonald, E. J. and Macdonald, F. A.: Z_va1uat?.:rn of Cancer Control Methodology. %:m J Public Health, 3u:(S,r 34. Macdonal ?atient Delaf. Co ;83-492, 1940. 21. Macdonald, E. J.: Statistical Researc7. C:,naac_i.:l: 35. .'•'acdcnal --Aa' ec] for Canc Division, Women's Field Arm of the American Societ^ of Cancer - er Bulletin, 52(3) 1 Control, Inc., Bu etin No. 1, 1942. , 22. Macdonald, E. J.: The Cancer Situation in Connect::- cut. Conn Health Bulletin, 56:(3), 1942. 36. :,acdonal Survival Picture - J N.-k Med Women's A 23. Macdonald, E. J. : Eight Years' Experience in Cz_ncer in T+enty-One Connecticut Hospitals. Conn State Med 37. '+.acc:onal. Rasults. Am J Roe: 536, 1943. 24. :dacdcnald, E. J.: Improve:.ent in Services ':?er.de.ad Conn 932-835, 39 . llr~- :•tacdonal: , in Connecticut Hospitals to Individuals with Cancer. Health Cu1let:n, 57:(8), 1943. Picture .. ?rcgno5is. 1iu11 A: 25. Macdonald, E. J:: Part Medical Record ;.ibrarians :•lay 29~-2>3 (12) 57 39. t•tacdonal, , : : in Disease Studies. Conn Health Bulletin, . . :e olugv oE 1943. aca_ r,my o: Sc' 26. '•t.uu:lun.7ld, i:. J. : Thu 4laf.,.-wicfc Cancer Ruwr:: 3t=,-s•:r1 40. Clark, R. in Connecticut. The }tedical 'rluman's J, 51:26-29, :5, 19:;. ~= Te:cas ;•1. D. Andc ^nen's J, 1949. 95-077 0-82-98

tistory of the Massac`:uset=s :e Commonhealth, 2: (2i, 43 Iistoricnl Trend - J ~b :abetcs Number, 24;2i, 1937. :uracy of the Cancar Dea':: 3:(7):818-824, 1938. + Activities of the ?a1th, Cancer Number, 3tory of the MassaC:au5,:_-_CS :h, Cancer Number, 2 Evolution of Cancer Cor.tro: i's J, 45:264-270, 1938. a Humanitarian Movement. 26:(3):176-179, 1939. •tacdonald, F. A.: Zva:ua:.i::n .:m J Public Health, 3u: (5; ; atistical Research. C:.nzec-i ~' the American Societ•• of C-nr•c•2r 1942. e Cancer Situation in Con.^.ect:`- (3), 1942. 3ght Years' Experier.ce in Cc_ ncer tals. Conn State Med - 7 (3; : orove:.:ent in Services 3er•cerac ividuals with Cancer. Co.^.: rt Medical Record Librarianz .1ay a Bullctin, 57:(12):29~-23:, .. Slat.-.-wi 3e Cancer Ruco::: nan's J, 51:26-29, 35, 1944. Publications (con't) -3- E1,.!anor ,,. 27. '•tacdonald, E. J. : A Report of Living Inci•: i"!::::Li t•ino !lavc ;l2ci Proved Cancer. Conn Health Bu11.2tin 53: (:; : 82-84, 1941. =8. Macdonald, E. J.: Connecticut :1s a Cocperativ•_: Entity for Control of Cancer. Conn Health Bulletin, 1944. 29. M3cdonald, E. J.: Survival Rates of Ind•iv?duuls i•tho Have Had Proved Cancer in Connecticut-1935-19•l2. :or••^. Health Bulletin, 58:(11):844-848, 1944. 30. Macdonald, E. J.: Expanded Chronic Disease Prc;ra:: Needed. Conn Health Bulletin, 60(8), 1946. 31. u;.cdonald, E. J.: The Meaning of Adjusted 't;:to.: for Cancer. Conn Health Bulletin, 60:279-282, 19;c. 32. Griswold, M. .H., and Macdonald, E. J.: The Connecticut Cancer Record Registry -How it Functions. State Med J 11: (5) :3•)4j347, 1947. - 33. Macdonald, E. J.: The Incidence of Cancer r_.._ 9reast Among Women. Conn Health Bulletin, 61:(7), ;3;7.' 34. Macdonald, E. J.: Chan:Ting Reasons for Cancer Patient Delay. Conn Health Bulletin, 61(4), 1947. 35. "acdenal3, E. J.: The Estimated 5umber of iuais Treate•.: _°or Cancer Needing Nursing Service. Ccnn He.:.it Bulletin, 52(3), 1948. 36. Macdonald, E. J.: The Present Incidence ar:3 Survival Picture in Cancer Among Females in Conr•ec__- :. J Am Med wcnen's Assoc,3:(4):152-162, 1949. 37. :Sacc:onald, E. J. : Criteria for Repcrtin c Results. 4n J Roentgenology and R3diu^:Theracv, 932-835, 1?48. 39. >12c'.onald, C. J.: Tho Pr,~scnt Incidance an,i Survi•:i! Pic*_ure in Cancer and thr~ l'romise of Improved ?rognosis. iiu11 Am Coll Sura, 's3:75-93, 1949. 39. ??acdonald, E. J.. Nalir:nruit Mclinoma in G.:r.r.•:c=icut. noas. Sf~ecia1 Publicatieas C. ,.. .. Iz The a:oicav of :4ela m crk acs~emv ot Sciences, 4:71-81, 19d8. - 40. Clark, R. L., and :•lacdonaid, E. J.: The _ of Texas '•1. D. Anderson Hospital for Cancer Research. Wo:nen's J, 1949. 96--077 0-82-38