Tobacco Documents Online

of a ghose th~se this ereas t~ed a ~k for ading r MI, onger imate ;ween barge ;iving .es in r cent )r MI ~, and S. We of the were dates • Mis- )osure pared Inder- might would in the h this, : esti- : more 5mate ~, were eports :d risk ~ssible influ- ques- whose ewed, ~g OCs eight rge di- i adis- : heart AND .~.~IYOCARDIAL I~'FAECTION 65 disease, and one (13 per cent) had a dis- charge diagnosis other than ischemic heart disea~=e. Considering that, overall, 81 per cent of the hospital summaries re- viewed had a discharge diagnosis of these sparse data do not lend support to the possibility of selective reporting bias. With regard to confounding bias, we controlled state of residence, calendar year, cigarette smoking, parental history of MI, and the presence of predisposing conditions. However, the iAability to con- trol severity of the predisposing condi- tions, which were associated negatively with OC use, might also have led to un- derestimation of the relative risk. In the present study, as in the previous case-control studies (1-6), past use of OCs was not associated with art increase in MI risk. Nor was there evidence of h consis- tent trend of increasing risk with increas- ing duration of use. The other case-control studies, particularly the most recent (6), in which the women had the longest mean duration of exposure to OCs, also give lit- tle indication of an association ofduration of use with MI risk. Among women in this study who did not smoke and who had no other iden- tiffed risk factors, current OC users ap- peared to have a risk of MI 2.8 times that of women who had never used OCs, which is consistent with the fourfold increase in MI risk estimated for such women from the data of Mann et al. (3). Regarding the joint effect of risk fac- tors, we observed that the greatest rela- tive increase and the greatest absolute in- crease in the risk of MI associated with OC use occurred when both cigarette smoking and hypertension were also present. Furthermore, the increase in the risk attributable to the three factors act- ing jointly appeared to be considerably greater than what would be predicted on the basis of the sum of the risks attribut- able to each factor separately (19, 20). This is consistent with Oliver's observa- tion that a disprsportionate number of women in his case series ofyoung women with MI had multiple coronary risk fac- tors (21}• Also, although they did not evaluate specific combinations of factors, Mann et al. (3) estimated that women with any three or more risk factors ap- peared to have a risk of MI 120 times that of women having none, an increase in risk greatly exceeding that expected from the effects of the factors acting alone• Re- cently, based on a larger number of obser- vations, Shapiro et al. (6) estimated the relative risk for current OC users to be similar among nonsmokers and heavy smokers; a constant relative risk implied a much larger absolute increase in the risk among heavy smokers because they were initially at considerably higher risk than the nonsmokers. Thus, our data, together with those of Oliver (21), Mann et al. (3) and Shapiro et al. (6), suggest that the increase in risk associated with OC use and other coro- nary risk factors differs according to the underlying level of MI risk, and is greater among those initially at higher risk. This is already reflected in OC package labels, which include cautions about the effects of pill use on the risk of MI in older women, in heavy smokers, and in women with other predisposing factors. REFERENCES 1. Mann JI, Vessey MP, Thorogood M, et ah Myocardial infarction in young ~vomen with special reference to oral contraceptive practice. Br Med J 2:241-245, 1975 2. Mann JI, Thorogood M, Waters WE, et ah Oral contraceptives and myocardial infarction in young women. A further report. Br Meal J 4:631-633, 1975 3. Mann Jl, Doll R, Thorogood M, et ah Risk fac- tors for myocardial infarction in young women. Br J Prey S~ Med 30:94-100, 1976 4. Mann JI, Inman WHW: Oral contraceptives and death from myocardial infarction. Br Med J 2:245-288, 1976 5. Mann JI, Inman WI~V, Thorogood M: Oral con- traceptive use in older women and fatal myocardial infarction. Br Mecl J 2:445-447, 1976 6. Shapiro S, Slone D, Rosenberg L, et ah Oral con- tmceptive use in relation to myocardial infarc- tion. Lancet 1:743-747, 1979 TI09781828

66 EO~EE~ ET AL. Beral V, Kay "CR, P~yal ColIes-~ of C~n~ral Pr~c~ti~ne~ ~ Csnt~pti~n S~y. l~or- ~ity ~ng cral ~nt~ve u~. 2:727-~31. 1977 8. V~ey ~, ~cPh~s~n ~ John~n B: M~lity among women ~rticipating in th~ Oxf~r~ Family Pl~ning ~s~iation Contraceptive ~ Study. ~ncet ~731-733, 1977 9. Kay C~ Oral Contrs~ptives and Health: An ~im ~rt from the Oral ~ntra~ptive Study at the ~ya] ColIege of ~nvral ~a~- ti~one~. New York and ~ndon, Pi~an, 1974 10. Very ~, ~H R, Pe~ ~ et ah A follow-up ~udy cf wcmen ~i~ diffe~nt vds of contraception--an inte~m re~. J Bi~ ~i 8:371-424, 1976 11. He~ekens CH, Ma£M~n B: Oral ~ntra~e~ rives and my~dial i~a~t~n. ~i~rial. N Engl J M~ 20:1166-1167, 1977 12. ~l N, Haens~l W: S~ti~tical as~c~ofthe a~ly~is ofda~ from retros~ctive studies ofdis- ~. J Nail C~r Inst 22:719-748, 1959 13. Miettinen ~: Simple inte~al e~timation of ~k ratio. ~ J Epidemiol ~00:515-5~6, 1974 14. ~m~ ~: ~mpu~tion of exact covalence i~rva~ for th~ c~ rat~. ~,t J Bi~,~.ed Cem- p~t 6:31-~, 1975 15. Walk~ SH, Duncan DB: E~timnti~n vf the ~ility cf ~ e~nt as ~ f~un ~f ~1 in~a~ent va~b]~. Bi~me~ka 54:167-179, 1~7 16. G~ JJ: ~e ~m~r~n Cf pmF~iens: A m- view of si~fic~ce ~, ~en~ in~'als, ~ adj~n~ for s~atifi~fion. ~v Int Stat Inst 39:148-169, 1971 17. Jick H, Dinah B, Rothman KJ: Oral con- ~aceptive~ and non-fatal myo~rdial infa~- tion in otherwise healthy women. JAMA ~9:1403-1~, 1978 18. Beral V: Cardiovascular disease m~rtality trends and oral contraceptive use in young wom~n. ~ncet 2:1~7-1052, 1976 19: ~e P, ~cMahon B: Attributable risk ~nt m case-~ntrol studies. Br J ~ev Soc Med 25:242-2~, 1971 20. Rothman KJ: Synergy and antagonism in cause-effe~ relationships. Am J Epidemiol 9~385-388, 1974 21. Oliver ~: I~emic h~rt disease in young women. Br Med J 4:253-259, 1974 an pr, tio hi~ ily S in i | cuh que thi~ T 8881 tim" effo tior Div T109781829

FOR F_..~XOEMIOLOGIC F~FJ~.CH: A~-~C'~S t~.e ~c~r.s tr~'r~w were th~ cnly t~o f--c~rs signifi- cantly as~wiated with l~ukemia. Pros~tic cancer in Utah, 1~6-19T7. D. West* J. Powell (U. ~U~, ~It ~e Ci~, ~ 1~ ~d 19~, 3140 c~ ~e p~ we~ ~ ~ ~ U~ C~cer ~- ~. ~ ~ y~, ~ in~denm in~ 359 ~oc.~hiz~=.l-~re~is (SADS-L), a s~ru~ured clini~l int~r- view dsve~,~'d to elid= SigT= a~d sy~p~:~s cf ~sy- chiatr~c diKur~a~ce. Ba~ on t~e ir~formation col- lated in the SAD~.L tl'~e ~Jbject~ were clcssifi~i cn the P~sear~ Diag~csti~ Criteria (RDC), a s~t cf cp erationat defmiticv~ with R~'¢i~c i~cl~ion and ex- chmicn criteria for a variety cf p~y~tr~c dL~rdem C~pyright ~ 1979 by ~e Joh~ Ho~ Unive~Ry ~ of Hy~m~ ~d ~b]~¢ Heal~ ABST~C~ OF PAPE~ P~SENTED AT THE ~EL~H ANNUAL MEETING OF THE Vol. 110. No. 3 Society for Epidemiologic Research New Haven, Connecticut June 13-15, 1979 R~ommendat~on~ for the anallmis of t~atment f~ on ~ de~l~nt of~nd m~i~. Makuch* ~d ~ S~on (Bio~c ~ B~, NCI, ~ ~ 2~14). U~ of ~e ~-ye~ ~n~pt ~ ~vi~ for eval~fing ~ ~a~on ~n ~t of a p~ma~ ~r ~ sequent ~val~t ~ a ~ond ~~. ~ implidt ~mpfi~ ~ this me~ ~ ~ sho~ ~me ~ inaction. ~fl~ly, ~ ~ of ~ meth~ 1~ W ~ ~vaHd ~ of ~k of developing s ~nd ~i~an~ w~ ~e ~enm follow-up ~H~ For ~m~ ~nfi~ of two or mo~ ~n~, ~ ~ bi~ ~avo~bly agai~ the ~ of in~ive ~at pmlo~ l~e. ~m~t~n ~, and two all--five p~, ea~ f~ ~ in a ~m~nly ~in ~e ~latio~p ~tw~n Weatmsnt velopment of a ~nd ~r. Potpourri Chairman: F. C~arimg Screening for depr~sion in the community. J. I¢_ Myers, M. ~L Weiesman and W.D. Thompson* , (Yale U., New Haven, CT 0~510). In 1975-1976, 515 p~r~orm dawn from a commu- nity survey in New Hav, v. Commcticut,.w~re admin- istex~ s~veral widely u~d self.report deprs~ion symptom scal~. In addition, they were interviewed using the Schedule for Affective Disorders a~d remote group of British island, into the home of Eur~po's lm*g~t oil ~ B~-au~ of a special par- liamentary bill, all of the oil development~ were re- stricted to one zone (the study's target area). In 1975, just before m~jor cormtru~tion had begun, the authors carri~l out a ba~line study with stratified ~amples dawn from National Health S~rvices rolls to reflect the age and sex characteristics of the target area and a control rt~ion (s cormervation area). The original ~amplo wm, mad~ up of 533 indi. viduais (targat, N = 263, control, N ffi 270)betwben the ag~ of 15 and 60 yeRrs. Tha beeline leveis ~ ~iotm or treated lmychiatric symptomatology were low: over~l level = 4%, depr~mive symptomatology = 5%. The fi~t follow-up study wa, completed in 1978 (peak ol'the construction poriod) and 94% of the original cohort was r~interviewed. This report con- cerns it,ll with changes in psychiatric symp. tomatetog~ met how them chm~g~ relate to alcohol, tobacco and l~ychom.'tive drug tmage. Cigurett~ smoking ~j~~m~l me~u~. ~ V. Kaufma~* D. Slo~ I~ Ro~aberg,~tO. Mie_ttinen '| and S. Shapiro {Dru~*~/~nit. Bo*ten, . MA 02138). ', Several ~t~lis~ have observed an inveres reiatio~- ship between cigarette imoking a~d age ~t natural menopaule. In order to further qu~nti£y the re- lationship, ~moking habits and age at menopause were compared in 656 hat.ally po~tmenopau~a! women who were 60-69 yearn of age, and who reached their menopat~e between the ages of 35 and 59. All of the wom~n who smoked had done ~o st lea~t since the age or" 35 ~e~ The mean age at menopa~ declined from 49A year~ among worsen who had never crooked, to 47.6 yesr~ among women int! a I~ 4O mi T109781830

whn snaked at least 15 c~gar~tte~ per c~y fp < w~ ~ey ~I~n~ hy ~e ex~n~, at ~e ~me cf thJl study ~m~ide qulntitat~ve evidence that c~aret~ crooking i~ ~s~a~d with an e~ly menopa~. A Icng-~ follow-up o( h~k~et~c ~d ~n- hy~r~netic ~tldren. ~. C. Howell* and H. ~ 1~ ~e~ fill~ out qu~on~ on ~ ~ s~en~ in ~w~ V~on~ ~a~ ~ ~e ~ h~etic ~me. exam~ a~r the 8tuden~ had ~mple~d ~de, ~ ve~ ~ d~ ~ ~~ ~ ~ ~e aurora have ~w ~ 1~ ~de ~ of ~¢ ~e in~d~ {'h~er~ic"} Jn ~e e~emen~ ~adel have Io~ ~e ~t ave~, am mo~ l~kel~ ~ ~ ~d ~ve a M~h~ ~t m~. It is ap~nt v~ for la~r ~avJgr. Mo~ver, ~uent work on ~e ~t~ of~e qu~t~o~ h~ e~a~n~ i~ ~ ~d ~t~flity of ~e ~e and ~ sho~ it ~ ~ ~ ex~mely ~ful ~1. C~nt work ~ ~ k ~nc~ ~ in~e~n~ ~e su~j~ ~ ye~ ~r ~ey have ~dua~ ~m A ~ ex~Hment of heal~ e~ of air ~llu- ~P ~n* ~. of N. Ca~lJna at Cha~l ~Q, r~p~a~ d~ (C~} in a natal setting. ~ ~ ~ ~, ~t Bi~n~ham had ~er ~e ~mdy ~p~on ~ ofpa~n~ of~il~ at~n~n~ elementa~ sch~l in thr~ selected ne~~ ~n ~th ~. ~o~ w~ • s~bu~ in Novem~r, 1~I. Qu~tio~ on were d~v~ f~m ~e 19~ B~t~h M~ ~ ~l'l CRD quahog. ~ paper fin~n~ on ~e 4~ whi~ ~ s~. C~ cl~ifi~ ~n~ five levels of ~v~ty. ~ we~ c~ ~or mal~ in ~ out of ~oups: non-job-exposed non-smoke~, non-job- exp~ed ex.~mzkers a~d.~x'~d current er~. Fer £e~, all ~ups exit ~ns, ~nt ~:~, sh~w~ ~ ~Hu~n eff~ Using ~e ~K m~ing me~, B~~ had ~gh~ e~ ~t val~, ind~ wc~ C~ ~n~, ~ ~ Ch~lo~ ~&n~. ~e ~e~nm w~ ~t sight f~r ~l~, hut it cl~ ~ s~t f~ f~. Smok~g ~d ~ ~n~ mu~ ~r d~e~n~ in C~ ~n ~d ~r ~lluti~n ~on~ No si~ifi~t ~rac~9~ we~ Blo~ lead level~ and non-~eciflc symptcnm in a smnple of smel~ emplo~ in Tmfl, B~fish ~l- umbia. L.C. N~, H. L. ~o~* a~ D. He~ ~y~ ~wa H~i~L ~ ~o, Ca~ KIN 6N5). BI~ l~d ~n~n~o~ {~B) ~m me~u~ in a ~[e ~ ~ Employs. and their w~v~ ~ T~l, ~W of one of ~e I~ smel~ in No~ America and in a ~arby cont~l ~wn ~NeI~n, ~ ~ ~eir ex~ ~ lead ~d ~e average ~B con~n~tion of 41, 33 ~d 16 ~dl which w~ fo~d refl~ ~eir work. ~ail w~en and Nelson con~l~ ave~g~ 11-1S ~g/dl. No geo- ~aphi~l ~t~ {by residence) w~ discernible a~ T~I women. ~e smokem had ~ifi~n~y higher av~ge ~B levek {~R ~) ~ non- worke~. ~ a~mpt ~ rela~ ~B ~ s~ms clio- i~ by qu~o~ ~ ~sc~ For ad~t ~l~, si~t ~a~o~ ~m fo~ for ~ven s~ ~, all ~ w~ were biolo~lly pla~ible- N~e co~la~ BI~ l~d w~ a~ me~ in a su~ ~mple ~ 80 work~ (and ~ek fami~) for each e~ ~up ~r a fo~-mon~ ~e ~d at I, 2 and 4 ~n~ ~ak ~r ~e plant ~um~ atlon. ~ hvel~ ~ e~ ~rk~ ~ch~ their m~m~ a~r one mon~. ~Ithough th~ ~ w~ n~ ~c i~ ~i~n~ w~ ~nfi~ by pair ~aly~i~. Relation ~tw~n multiple sclems~ and den~l ~es. W. Cmeli~ (~ayet~ C~lege, Eu~n. PA S~ng p~e~ ~n ~e ep~de~ol~ of multi- ple ~e~is (~) ~d ~n~ ~ ~ reve~ed ~mp~ng ~ ~m publ~h~ ~onal dental-and MS su~eys. ~e prevalence of ~ in ~lat~ well ~ chics inci~ ~ 0.~IL In A~lian states, dea~ ~s ~e linearly ~la~ ~ inddence of~es ~ ~ 0.97,p < 0.~2). In • e U~t~ S~ a ~sitive co~lation - 0.65,p < 0.001). Ja~e have ~ low MS an~ low ~ m~. No,hem Ireland ~d the ~ott~h ~I~ ~k h~t in ~ MS a~ dental ~se m~. ~n~ heal~ ~ ~n mpi~y d~l~ing in ~h kl~, ~leling ~e mpi~y ~ing rate of~. Co~la~ng ~e ~n~ of ~en~o~ in- • vld~k ~ ~ ~len~ in ~ no~ern ~e ~sk for MS and den~ ~ ~ lower amon~ bla~ ~mpa~ ~ w~, m~ ~mp~d with in the U: that ccm pre~sl~s. The excg summa" P~ka~. ~Messa~ 02114~. ~e • e Orkm tic a~ly~ stated a a pr~om~ ~ o~c~ did not mo~aphi, and ~sley mov~enl sho~ though in~ l~iy a ~tien~. year ~d lifetime all subj~ s~ti~ ~in~ in a~ ~e tim. but two e~ etiolo~ of lifetime of Polissa~ ~r, ~attle A numbt risk ~om ic~ indusr ~ cancer ~te and ~d~ou weake~ of~o~b and ~ex~ lat~ the ef esti~t~ c, the t~e ~sk minus relation b mat~ by 5 morality d: Tables estimation ~inciden~ TI09781831

ruptured sac of cerebral aneurysm must be assumed to be sealed by a plalclet plug which gradually organises into a blood-clotY1~-~9 Blood in the subarachnoid space stimulates fibrinolytic activity. This is shown by an in- creased amount of fibrin-degradation products in the cercbrospinal fluid (C.S.F.) of these patientsn.~" and by a raised level of plasmin in plasma and C.S.F.'° In som~ patients this fibrinolytic activity leads to lysis of the pro- tective clot and a recurrent ha:morrhageY~.~ Fibrinoly- tic activity gradually increases during the 2nd and 3rd weeks after the hmmorrhage, and this accounts for the incrcascd incidence of recurrent ha:morrhagc during this period. z.^.c.^, acts chiefly by competitive inhibition of the activator that converts inactive plasminogen into active plasmin, a proteolytic enzyme}~.11 L.~,.C.A. also inhibits the effect of plasmin2-~ and reduces the blood-plas- minogen levelY This pharmacological effect of ~.^.c.^. helps to prevent lysis of the protective blood-clot and leads to improved clot maturation,s.17,~z thus preventing recurrent ha~morrhage. Gibbs and O'Gorman1 in 1967 found no conclusive evidence that ~:.^.c.^. alters the prognosis in. S.^.H., whereas Mullah and Dalwey in 1968 found that e.A.C.^. sharply reduces the rate of recurrent hmmorrhage. Only four studies~'J.x~. used a control series, and three of these''t-~ did not reveal significant reduction in recur- . rent h,'emorrhage in S.A.n. This prompted us to under- take the present study. Our trial was not strictly randomlsed, since patients were allocated to the two groups according to the day of admission. Definitive surgical management differed. somewhat in the two groups, and hence we have limited our report to consideration of recurrent hzmorrhage and to mortality before surgery. We did not encounter an increased incidence of blood coagulability or deep-vein thrombosis. This accords with the results of Nilsson et al.,~s Sengupta et al.,~ and Nib- bclink)~ We did not see any arteriopathic complications in the cerebral vessels as reported by Sonmag and Stein.~ Drug-related nausea and vomiting affected 10~ of patients, and 3 patients (4qe) temporarily had to be given the drug intravenously to control diarrhea. Since the immediate-recurrence rate was reduced to 4.'7; with administration of ~.^.c.^., it can he said that the drug improves the natural history of s.^.tl. We recommend that r..^.c.^, should be used by the primary- care physician as soon as a diagnosis of s.^.tt, has been made. This will greatly reduce the rates of recurrent ha:morrhagc and mortalily (hot'occur between the diag- nosis ofs.A.H, and the patient's transfer to a ncurosurgi- cal unit. Only 1 of 82] patients receiving v_.^.c.^, died from recurrent harmorrhage, and with the passage of time most of the patients showed apprcclable clinical provement. 97~.~ of the patients who~e s.^.H, was caused by cerebral-artery aneurysm were operated upon.~ Even 743 SAMUEl DENNIS SLONE DAVID \'~:. K^UFMAN Drug EpidemioloILv Unit, Boston Universio, School of 3dcdicine PAUL D. SZOLLEY. Dcparonent of Research Mcdlcinc, OniverH~v of Penn%vh~nla School of Medicine OLLt S. MIETTINEN Department of l~pide.dolo~v. Harrard Sclwol t~Public lh'ahh Summar.y The effect nf oral-contraceptive use on the risk of nlyocardial infarction and, in particular, the possible accentuation 0f that clrcct by cigarcue smnking, was investigated in 234 prcnlcnO- pausai women with a lirst infarction and 1742 hospital controls. The overall rate ratio estimale of acute myo- cardial infarction for women who had used oral contra- ceptives in the preceding month was 4-0 (95~.~ confi- dence interval, 2.5-6-3). Women who smoked heavily and used oral contraceptives had a point estimate of 39 (lower two-sided 95¢[ confidence limit, 22) compared with thane who did neither. This value was appreciably T!09781832

ruptured sac of cerebral ancu~,sm must l~ assum~l to be s~al~l by a pIatel~t plug which gradually organises into a blood-clot.LlT-l~ B|ocs:l in the subarachnoid space stimulates fibrino!ytic activity. This is shown by an in- creased amount of fibrin-degrad'ation products in the cerebrospinal fluid (C.S.F.) of thee patientslz'l* and by a raised level ofplasmln in plasma and c.s.~)a In some patients this fibrinolytic activity leads to lysis of the pro- tective clot and a recurrent h:-morrhage.~-*.~ Fibrinoly- tic activity gradually increases during the 2nd and 3rd weeks after the hmmorrhage, and this accounts for the increased incidence of recurrent hmmorrhagc during this s.^.c.A, aas chiefly by competitive inhibition of the activator that converts inactive plasminogcn into active plasmin, a protcolyzic enzyme.~'~ ~.^.c.A. also ~nhibits the cffc~ of plasmin~ and reduces the blood-plas- minogcn loyal.~" This pharma.cologieal cffcc~ of ~.^.c.A. helps to prevent lysis of the protective blood-clot and leads to irnprovcd clot maturation,~'xv'~ thus prcvcntlng rccu~cnt h~morrhagc. Gibbs and O'Gorrnan~ in 1967 found no conclusive cvidcn~c that ~.a.c.A. ahers the prognosis in-$.^.H., whereas Mullan and Dalwcy in 1968 found that E.A.C.A. sharply reduces the rate of recurrent h~-morrhagc. Only four srndies~,~.x*. used a control scrles~ and three of these~-~-~ did not rcvcal significant rcduaion in recur- rent hzmorrhage in s.^.IL This prompted us to under- take the present study. Our wial was not strictly randomiscd, sinc~ patients wcrc allocated to the two groups according to the day of admission. Definitive surgical management differed somewhat in the two ~'oups, and hence we have limited our report to consideration of rccurrcnt h~'morrhagc and to mortality before surgery. We did not cncounter an increased incidence of blood coagulability or deep-vein thrombosis. This accords with the resuhs of Nilsson ctal.,u Scngupta ctaI.,~ and Nib- bclink)~ We did not see any arteriopatMc complications in th'c cr_rebral vessels as reported by Sonntag and Stcin3~ Drug-related nausea and vomiting affected 10% of patients, and 3 patients (4%) temporarily had to be given the drug intravenousIy to control dlarrh~a. Sincc the immediate-recurrence rate was reduced to 4% with administration of s.~.c.A., it can bc said that the drug improves the natural history of s.~.~. ~c recommend that s.a.c.A, should bc used by the primary- care physician as soon as a diagnosis of s.~..~, has been made. This will greatly reduce the rates of recurrent hzmorrhagc and mortality that occur between the diag- nosis of s.A.~, and the patient's transfer to a neurosurgi- ca| unit. Only 1 of 83 patients rcccixdng ~.~c.A. died from recurrent h~morrhagc, and wi~h the passage of dmc most of the patients showed appreciable clinical provemcnt. 9~% oft.he patients whose S.^.H. was caused by cerebral-artery ancurysm were o~ratcd upon3~ Even gro~ly cbmn:L'd pa~/~ms irr~ruved so mu~ ~at su~l mo~afi~ ~d m~dixy ~uld ke r~onably ~ to ~low.~-~a 1.~ W, W~, L Br. ~.J. 1956, 7~ 559. 2. Oi~ J. R. 0"~% P. Po#~=J. med.]. 19~L 4~ 779. 4. 5Mw, M. D. M., ~, J. D. ~:a, 1974~ ~t 84X ~. Utah, D., R~Fm~ & E. i~'d. p. lOgO. 6. P~ ~. Neuror.~. 1977, ~ ~. 9. Mul~an, S~ ~w~,J. F.ibid. 1968,~,21. 12. Ni~link. O. W. ~bmva~]ar ~p. 155.N~ YorE, 1975. 16. Hunt, W. E.. H~ R. M.~. Ileuma~r~. 196S~ 2~ 14. I ?. Mtu~WilIiam~ R. S. Br. med.J. )97~, ~ 945. I 8. To~ D. Acta .e~. ~cand. 197J, 4~ 163. 19. ~mpt~, M. R.~.N¢~oL~euro~ur~. P%¢~M~t. 196~,29, ~. 20. ~k~ H. B.J.Ne~r¢. ~6,~, 32L 2L Alk}~g, N.. ~ch~, ~ P.Sh~, S.J. bill Chem. 1959~ Z34s 1959,~ 15~. 2J. ~tish Patent s~fi~tion, ~93. H. M. Statton~ O~ce, 19~Z 24. Nib~Iink, D. ~., Ja~ C D. Thromb. Diarh. ~mo~. 1973. 29, ~98. 25. Nil~, I. M., A~n, L, Bjorkman, S. E. Acta me6 stand. 1966. ~ 26. ~nnta~ V. K. H, 31~n~ B. M.J. Neuro~ur~. 1974, 4~ 48. 27. ~o~ha~, U. M., ~ P. C., Hu~in~ M. M. 29. Drake, ~ G. ibid. 196B~ 2~ ORAL-CONTRACEPTIVE USE IN RELATION TO MYOCARDIAL ]~FARCTION Medfcine P^U~- D. SVO~L~Y Depar¢mem of Re~earch Medicine, Unieer~iry of Pennsylvania School of Medidne Department of Ep~dem~olo~,, Ha~ard School of Publlc Heal~h Summa~ The ¢ff~t of oral~ontraceptive use on the ~sk of myocardial infarction and, in pa~icular, the ~ssible accentuation of lhat effect by cigarette smoking, was inv~tigat~ in 234 premen~ pausal women with a first infa~tion and 1742 hospital cont~l~ ~e ov~all ~t¢ ratio ~timate of acut~ my~ cardial infarction for women who had us~ oral contra- ~ptives in the pr~edlng month was 4.0 (95% denc~ int~'al, 2-5-6-3). Women who smoked heavily and used o~! contraceptiv~ had a ~int ~tlmate of 39 0ower tw~sided 95~ c~d~ce fi~t, 22) compared wilh lho~ who did neither. ~ value was appr~ably T109781833

MYCC.~RI}I.~J. INFXHf~I'[ON .XND CIC;,:~,RE'FFE-e. - ~LONE, ET AL i~75 k~m~ Fo~th=~n #asmm Clin Chim A~a 63:33G24~. 19"/5 Gmt~ H. Laiblc V, ~p~rIc G. ct aE Comparison of a~a? melhod~ for ~Fafi* ~d tip,protein l~pa~ in normaI~ a~d patients with hy~tri- gly~rid~mla. Ate:reset:rests 26:55~572, 1977 trigly~rid= [ip~ m uremic patienl~. N Engl J Med 297.l~52-i3~5. 1977 Clan ~ 8:63~7. F~man M. Ku~kcn ~ R~gland In, ct ~I: H~ati= ~I)zcdl: 35 De Vnm O. Vo~ J. ReiBma WD. eral: Muscutar ~pilIa~ ;+te~l memhran~ hl~rophy in h}~tlipaemll DFe IV: a con~qucncc uf a ma~kol~c dis~rd:r? Ncth J Mcd 17:~226. 1974 3& Vital C. ke Blanc M, Vallat JM. ctal: Etude uhrastru~ural= da n*rf diab~tiq~es. Acta N~uropalhd [Betl] 3D:o3-72, 1974 RELATION OF CIGARETTE SMOKING TO MYOCARDIAL D~:N.Xts SLOr~E, M.D, SA:aUE~- StIAPtaO, M.B., F.R.C.P. D~,VlD W. KAUFMAN, B.A., STUART C. HARTZ, Sc.D., ALLEN C. ROSSI, ArqD OLLt S. MIE'rrtNE,% M.D. Abstract To examine the relation between myocar- dial infarction and cigarette smoking in young women, we investigated the smoking habits of women under the age of 50 who had survived a recent myo- cardial infarction. They had not been using oral con- traceptives, and other identifiable risk factors were ex- cluded. Among 55 such women and 220 controls matched for age and area of residence, the propor- tions of cigarette smokers were 89 per cent and 55 CIGARETTE smoking increases the risk of myo- cardial infarction. This relation is most evident at the youngest ages and becomes progressively weaker with increasing age. Although most of the evidence has been derived from studies in men,1'2 Mann et al. have shown that cigarette smoking is a strong risk factor for premature myocardial infarc- tion in women as welP; women at high risk were in- eluded in their study. We are conducting an extensive case-control study to examine a large number of risk factors related to myocardial infarction in women below the age of 50 years. In this report we examine the relation of risk of myocardial infarction to cigarette smoking among women who are otherwise apparently in good health and who are not using oral contraceptives. SUBJECTS AND METHODS From.July, 1076, over a period of about 1 U.. years 1"~2 hospRals having coronary-care units agreed to collaborate m the ~,tudy. Nurses were trained to conduct interviews using standard ~orms. and then stationed in Boston and its surroue.dings (6fl hospitals), Long Island and the coastal area north of Nev, York City (48) and the Delaware Valley including Philadelphm and Wilmington (44). Telephone contact from a central office v, as made with each cot- chad'-care unit at intervals of eight to 12 days to determine whether any woman 49 years of age or less had been admitted. The physi- From the Drug Epidemi~logv Unit. Beseem Uni~erslt} Medical Center, the Department of Research Medicine. Ul'~iverstI} of Pcnn~)lv;mia School of Medicm¢, and the Department of Epidcmiutogy, Harvard School of Public Unit, 10 Me,hen St.. Cambridge. MA 0213~L Supported hy a contract IN01-HD-~2849) ~tth th~ National Institute of Child Health and Haman ~elopment. a contm~'t (22J-7~3~116} ~ith the F~ and Drag Administration a~d c~ntrzct~ (N~I-CP-T[029. N01-CB- 74~9) ~i~h the N~ti~nal C~r l~tute a~ ~ gr~nt from H~ff- ma~n-LaR~h:. 1~:. YOUNG WOMEN M.D., per cent respectively (P<O.001). A dose-response rela- tion was evident; among women smoking 35 or more cigarettes per day the rate of myocardial infarction was estimated to be some 20-fold higher than among those who had never smoked. This study dem- onstrates that cigarette smoking is a risk factor for myocardial infarction in young women who are otherwise apparently heaRhy. (N Engl J Med 298:1273- 1276, 1978) clan was asked if he believed the patient had definite, probable or possible myocardial infarction, ischemic heart disease but not myo- cardial infarction, or some other diagnosis. Whenever the diagnosis referred to ischemic heart disease in any way, permission for an in- terview was requested If there was any susIficion of myocardial infarction a nurse inter- viewed the patient, either in the hospital (83 per cent) or at home (17 per cent). Other patients in coronary-care units (ischemic heart disease, but myocardial infarction ruled out) were interviewed only if they had not yet been discharged. As potential controls, the nurses interviewed five or more women, as close in age as possible to each case, from the surgical, orthopedic and medical services of the same hospital, l)etailed medical histories (particularly regarding risk indicators for myocardial infarction) were obtained, as were comprehensive histories of any drug use at any time before admis- sion. Whereas data collection is continuing, the present report is based on data to hand as of December 31, 1977. Initial case series. Of 5~1 telephone notifications of definite or snspected myocardial infarction, or of iscbemic heart disease, from the units permission for interview was not given by the physician or patient in 3~t instances (7 per cent)+ Myocardial infarction was not considered to be the diagnosis in 337 women (256 of whom had already been discharged and were nut interviewed). In the remain- ing 176 the diagnosis had been assigned by their attending ph.vsi- clans. Of these, we excluded six for the following reasons: diagnos- tic data we~ not in confornlity with World liealth O~rganization t riteria~ Ifour women); an aortic-vah.e prosthesis ,~,as in place (one wnman); and a sickle-cell crisis had been followed by m~,ocardial infarction nn tile t4th ht~spital day f,+ne woman) -- leaving 17t] cases. lmtla~ contr~t sen¢,. A h~tal of 2942 women ¢,ere approached as puteutial controls; 6 per t rut refu~ed tb.e interview. Of 2775 inter- yielded 1324 were sclettrd for further consideration because they had first dlaq|to~;es iutbzed, a priori. 1,1 be urre!ated to clgarellt l"urdter e.vqu+vms..%ivie the porp+.,se of lbi'~ study was In ex,,+min,r the relatltn~ between slne~kin~ ,tlld rnw×iwdi:d infarrtiun in ,,,,'omen who i~thPrwise previ~ms m:o:atdiat t~da~cthm 114 t ,Is~. 20 controls); dr,ag-treated T109781834

t2-4 Table 1, Cigarette Smoking among the 843 Controta. According to Diagnosis and GeOgraphic Area. Tt~m~ I ~ 6U ~4 ~ ~2 19 14 A~dominal conditions I b3 ~0 43 h7 41 26 ] 6 Other conditions 352 163 46 129 3~ 60 17 Bost~ (;I I ~ 47 ~3 ~6 1~ PhiIad~tph~ IU3 42 41 4s 4~ 13 13 angina peetoris (22 and 4¢,), abnormal bh,od lipids (33 and drug-treated hyperlension (55 and 143); dru~-t~eated diabetes mel- litus (23 and 29); drug-treated ohe,qty {44 and 208); use of digitalig glycosides (four and 28) and use uf oral cnntraccptives within the mnnth before admission /13 and 93) (Man~ of the pallents, par- t*cularly cases, had more than one reason toe exclusion.) Patiems were also excluded if d~cy smoked an unknown number of cigarettes (none and nine). The exclusions reduced ~he series to 5S cases and 843 controls. On the basis of a review of dala ab~tra~led from the hos- pital records, all 55 cascs finally ~neluded in this report met the World tlcahh Organization criteria h,r "'definite my~ardial in- Table 1 gi,'cs ~hc distributions according to oiagnosis and smok- ing stalus among the controls. Musculoskele~al c~mdhions ¢prc- dominantly orthopcdk I and msuma accoumed fi~r 39 per ecru, ~as- troimesdnal disorders for 19 per eem, and a wide range of condi- tions for the remaining 42 per cent nf [he series. Ra~es of <~garette smoking showed reasonable ,~rcemem amo~g d~e dia~nu'.~dc categories. Although ~mokmg rule~ were also quite shnilar among the geographic areas, tl~e proportion of wunn'n >mohin~ los, than cigarcucs per day in file Philadelphia region (4~ per cenO M~hcr ~han ~hat in Boston i37 per cent), ~iHa Nc~ York bcing in- termediate (43 per cent) To ensure that the comparisons wt~uld nol be confounded, e,~ch case of myocardial infarction was malchcd to four controls from the same five-year age group, and from the same hospital O1 per cent or failing IhaL from the same area of rcsidcm~ (59 per cent). For three who were 45, 42 and 41 years of age 11 age-matched controls (from th~ same region) could not bc found; they were drawn inqcad from the 3S-year ~o 39-year age group since smoking sales did not vary appreciably" between the ages of 35 and 49 years. ~able 2 <ompares d~e 55 cases and ~heir 220 controls in term~ c~hnic ~roup, rcligion, marital s~atus, parity, menopausal status and years of education. These variables were similar in the two In the analyses that follow, slralificalion of Ihe da~a accordin~ categories of the matched variables, age am] area of residence, does nm materially affect the relative-risk estimates. R~u~s Of the cases, 89 per cent were smokers compared with 55 per cent of the controls (estimated relative risk = 6.8; X~ = 22.2; P<0.001). With women who had never smoked as the reference category (relative risk set at 1.0), the estimate of relative risk for ex- smokers (women who had nol smoked for at least one year) is 1.4 (Table 3). The corresponding estimates for the categories, one to 14, 15 to 24, 25 to 34 and 35 or more cigarett~ per day are 4 4, -1.6. lq and 21, respectively. Figure I shows that these estimate, on the natural logarithmic scaIe, are consistent with a linear r~re~sion model (Ln RR = ~X) (~oodn~s-of- 5t X~ = 1.9): ~e trend of incr~ing ~ladve risk with increasing cigarette consumption is statistically significant (P<0.001). From the model: the point es- timate of relative risk for smokers of 35 or more cigarettes per day is 22 with a lower two-sided 95 per cent confidence limit of 7. Coffee consumption was positively related to cigarette smoking. However, when this factor was tontrolled, it did not materially change the relative- risk estimates. Age and menopausal ~tat,s. Forty-four per cent of the patients ~ere 45 to 49, 32 per cent 40 to 44, and 24 per cent below the age of 40 years. Median ages in the cases and controls were 45 and 44 years, respective- ly. With unstable rates the association was somewhat stronger anaong younger women, but it did not vary appreciably according to menopausal status (Table 4~. Area of reddence. The relative-risk estimates were similar in the three geographic ;areas. Duration o_f cigarette smoking. Among the smokers the age at which they began to smoke showed little variabilily, and the median in the cases (19 years) was only slightly lower than in the controls (20 years). Therefore, we could not evaluate the possible in- fluence of duration of smoking. Proportion of myocardial infarctions attributable to cigarette .,m,~ing (et~,dr.:~ic/mctwn}. From the data in "Fable 4, 76 per cent of infarctions in women under the age of 50, thought not to be at high risk, are estimated to be at- tributable to smoking,v Table 2. Distribution of 55 Cases of Myocardial Infarction and 220 Controls According to Various Characteristics. no % no. Bl~k 6 l I 25 I l Protestant 17 31 ~ Cathohc 33 ~ 128 /~ish 3 6 13 6 Single 5 9 21 Nutlzpart:as I 3 24 ~4 25 Ptcmcnop.s~.: 3~ 71 143 6 Edz~n ~ 12 yr 15 27 ~ 33 TI09781835

127~ N=v=r smoked 4 7 7.t ~13 I]6" F_..'t-srnokcr 2 4 27 | 2 I 4 1-14 t~ [5 ~,t t'~ 44 15-24 I 25-34 12 22 ~'35 14 25 12 5, 21 DISCUSSION Myocardial infarction is a rare disease in young womcn~ and it is hardly surprisin~ dm~ dw mo.~ dear- cut evidence that smokin~ is a ink factor a! an c~rly age has been documented in men ~,: In billow-up studies that generally recruited stthjects who at the beginning were in a reasonable state of health, the age-adjusted rates of myocardial infarction among heavy smokers, relative to non-smokers, are of the 1000 70.0 600 ~0 200 5,0 aD t • OBSERVED RR ESTtMATE ] ~. ESTIMATE OF RR I DERIVED FROM MODEL ~ 0 95%CONFIDENCE LIMITS [ DERIVED FROM MODEl. • / tn(rR),O,619X / / / If" / I O t 2 _a.__ __ a_._____-J.______ NEVER EK-SMOKER 1 M 15-24 25-34 _>_35 SMOKED IX:l) (x=2) !x~3} (x:4] (x:SI CIGARET~E.'~ PER DAY Figure 1. R~lation of R%~atWe Risk of Myocardial Infarction to Cigarette Smoking. order of three!old for men bet::een the ages of ak:out 4~~ and 65 D-ar~.s~ However. the assoziation tends to [w t ,msit]cralfly ~tronger in the )t,ungest men. For ex- ample, n~.~, rv~czitly I)oll ~md l'ctot rcpnrted that Mier a 20-}~r l~llow-up obse~'ation of male British doctor, 5q had died of ischemic hea~ disease before the age of 45: the relati~e risks for li#t (one to 14 cigarettes per day~. intermediate (15 to 24), and hea~T smokers (2~ or more) were 7. 9, and 15 -- results in accord with our findings. Although the patients in the present study survived their heart attacks, Spain et al. have m'pt~rted that in ~,men, sudden death from coronary heart disease is strongl} zelated to cigarette smokingY Table 4. Relation of Myocardial Infarction to Smoking Habits According to Age and Menopausal Status. Agct~,r) <45 Nc~cr smoked 2 1-24 >25 15 17 25 Never smoked 4 44 1" 1-24 10 4l 3 ~'25 11 11 Never smoked 4 b8 I ° 1-24 16 60 5 >25 19 20 Never Sl|lt)kcd 2 32 I * 1-24 7 32 4 ,'25 "~ 8 14 * Rcrercn.e Our findings also confirm and extend those report- ed by Mann et al,, who studied 77 surviving women with myocardial infarction who were below the age of 45 years? These workers estimated the relative risk to he 4.0 fur smokers of 15 or more cigarettes per day after simuhaneous adjustment tbr the influence of hy- pertension, pre-eclamptic toxemia and oral-contra- ceptive use -- and the relative risks rose steadily as the amount of cigarette smoking increased. Their series was too limited in size to evaluate the relation cf smoking to myocardial infarction in the-absence of other indications of high risk. fOur series of 170 women with myocardial infarction contained a sub- t~roup of 55 below the age of 50 who were not discern- ibly at high risk. Comparison of these women with suitahle controls indicated that cigarette smoking, ,done, seemed progressively to increase the rate of myocardial infarction up to as much as some 20-fold ,throng ~,nmcn smoking 35 or more cigarettes per day. Within the age range of 20 to 49 years stipulated in tiffs study, there was a suggestion that the associatiun was strongc~ in .',oungcr women, but this effect T109781836

f'HE NE~,~, F_.NG1...~ND jOURNAL (~1- MEDICINE could have been due to sampling variation, tIowever. the finding is in line with the way in which age modifies the smoking effect in subjects over the age of 49 years: in most studies, the association decreases in strength until it ceases to be detectable at advanced ages. Menopausal status in our data also appeared to be without effect either as a risk factor itself or as a modifier of the effect of smoking. Contradictory findings concerning menopausal status as a risk factor have been reported2 Part of the. smoking relation may be accounted for by a tendency of women with certain personalities to smoke, but increases in risk of the magnitude reported here have not been attributed to any specific personality type.t° Similarly, it is possible that the heaviest smokers engage in the least physical activity; again, the reduction in risk attributable to exercise" would not explain the association. Although we did not determine lipid levels, Truswell and Mann have reported that cigarette smoking and serum cholester- ol levels appear to be unrelated,n Cigarette smokers dying of conditions other than coronary heart disease have been found at autopsy to have intimal thickening of the coronary arteries and an excess of atheroma,t~ This finding suggests that a duration effect of smoking in relation to myocardial infarction might be anticipated. In this study we could not evaluate the influence of duration of smoking. However, apart from any atherogenic consequences, it seems that cigarette smoking may exert a precipitat- ing effect since the risk was hardly increased at all in ex-smokers. When an illness is exceedingly rare, even relative- risk estimates of the magnitude reported here are of limited public-heahh consequence, and in the United States, mortality from coronary disease in ,,'omen below the age of 30 is negligible. By the age of 40, however, it approximates 10 per 100,000 per year,t* and by 45 years, it has approximately doubled again,t~ In addition, about half the women sur-,,ive,~ giving estimated total incidence rates of the order of 20 and 40 per 100,000 per year, respectively. From our data, perhaps a third of the patients with myocardial infarction have no reason to believe that they are at high risk; our findings suggest that in this category, some 75 per cent of infarctions could be avoided if women did not smoke. In addition, the findings of Mann et al.~ suggest that a substantial propor- tion could be avoided even when the risk is discern- ibly increased by the presence of other predisposing factors. Among American ,,omen belo~" the age of 25, and among teen-age girls, rates of cigarette smoking have been increasing.'*.t' Unless this pattern changes, the contribution of cigarette smoking to the occurrence of premature myocardial infarction in otherwise appar- ently healthy women will probably increase. This study represents the teumv, ork of the nurses, staffand physi- cians in 152 hospitals located in the northeastern part of the United States. with,mr ",~hose help anti ,upp*~rt it could nt~t have been done. 5pe, r,al t redlt is due to Cartd Enseki. the program co-ordinator, and the nurse~ responsible for arranging and conducting tlxe intervie~,s. Linda Paradis, Geraldine Christie, Susan Oboe, Dorothy Gray, Carol Palmer, Mar.~grat'e Ba, ber and Margaret Imbro. ~,~,¢ are in- debted to Patricia l'~attv.~,'ler. Marguerite Angelarti and Leonard (;art;rap fur assistance and to [}r 1). ~V. Hosmer h~r advice in the irltai'r.'~ls t,t this sludv REFERENCES 1.Doll R, Pcto R: Mortality in relation to smoking: 20 years" observa- lions on attic Brittsh doctors. Br Mcd J 2:1525-1536, 1976 2. Doyle JT, Da',~ber TR, Kaanel WB. et al: The relationship of cigarette smoking to coronary heart disease: the second report of the combined experience ut the Albany, NY and Framingham. Mass, studies. JAMA 190:886-890, 1964 3. Mann JI, Doll R, Thorogoud M, ct al: Risk f~etors for myocardial in- farction in young women. Br J Prey Sue Med 30:94-100, 1976 4. Ischaemic Heart Disease Registers: Report of the Fifth Working Group. Copenhagen. World Health Organizatzon, 1971 5. Grizzle JE, Starmer CF, Koch GG: Analysis of categorical data by lin. ear models. Biometrics 25:489-504, 1969 6. Hosmer DW, Hartz SC: Linear models for functions of the odds ratios in a 2 × C contingency table. (Biostatistics Technical Report 77-18). Amherst. Massachusetts, School of Health Sciences, University of M as- saehusetts, 1978 7. Miettinen US: Proportion uf disease unused or prevented by a given ex. posure, trait or intervention. Am J Epidemiol 99:325-332, 1974 8. Spain DM. Siegel H. Bradcss VA: Women smokers and sudden death: the relationship of cigarette smoking to coronary disease. JAMA 224:1005.1007. 1973 9. Coronary heart-disease in young women. Lancet .2:282-283. 1977 10. Jenkins CD: Recent evidence supporting psychologic and social risk factors for coronary disease. N Engl J Med 294:987-994, 1033-1038, 1976 1 h Paffeabargcr RS Jr. Hale WE: Work activity and coronary heart mor- tality. N Eugl J Med 292:545-550, 1975 12. Truswdl AS, Mann Jl: Epidemiology of serum hpias in soulhern Africa. Atheroscierosis 16:15-20, 1972 13. Auerbach O, Carter HW. Garfinkel L. ct al: Smoker's heart. Arch Intern Med 137:435, 1977 14. Mori)'ama ] M. Krueger DE, Stamler J: Cardiovascular Diseases in the United States. Cambrtdge. Massachusetts, Howard Untversit~ Press, 1971 15. United States Department of Health. Education. and Welfare. Vital Statistics of the United States. Part B (DHEW Publication No. [HRA] 76-115). Rockville. Maryland, National Center for Education Statis. t~cs. 1974 16. idem: Teenage Smoking:. National pattern of cigarette smoking, ages 12 ,broach 18. in 1972 and 1974 IDHEW Pablieation No. [NIH] 76-131). Bethesda. Maq,-land, National Institute of Occupational Safer)" and Health, 1976 17. Idem Adult Use of Tobacco )n 1975. Atlanta, Center for Control, 1976 TI09781837