PART I Current Information on the Health Consequences of Smoking Highlights of The Report LIdtlitioi~nl physiological and epiclen~iologiral evidence confirms the previous findings that cigarette smoking is the most important cause of chronic non-neoplnstic hroncliopulmoiinryv disease in the United States. Cigarette smoking can adversely affect pulmonary function and disturb cxrdioplllnlollarv ~~h~siolog~. It is suggested that tllis can lead to cardiopuln~on:~rg disease, notably pulnlon:lry ]lypertension xnd car pulmonale in those intlividuals who 1Mve se\-ere chronic obstructive bronchitis. 3 SmoA-ing and Cancer Additional rvidcnce substantiates the prcrious finclings that ciga- rette smoking is the main cause of lung cancer in men. Cigarette smok- ing is causnll\- related to lung cancer in n-omen but accounts for a smaller proportion of cases than in men. Smoking is a significant factor in the causation of cancer of tile larynx and in the development of can- cer of the oral cavity. Further ep*idemiological data strengthen the association of cigarette smoking with cancer of the bladder and cancer of the pancreas. Smoking and Overall Mortality The 1964 Advisory Committee's Report (3) clearly and em- phatically outlined the dangers of cigarette smoking to health. The conclusions of the Committee, as outlined in the 1067 Report (2)) were as follows! CIGYYRETTE smoking is associated wit11 n `i&percent increase in the age-specific death rates of n~;~lcs. and to a lew'r extent with in- creased death rate of females. The total number of excess deaths causally related to cigarette smoking in the l'.S. population cannot be accurately estimated. Tn view of the continuing and mounting evidence from many sources, it. is the judgment of the Committee that cigarette smoking contribnte9 sul)stantinllv to mortality from certain specific diseases and to the overall tleath'rate. In general, the greater the. number of cigarettes smoked daily, the higher the, death rate. For men who smoke fewer than 10 cigarettes a day, according to the seven prospectiT-e studies, the death rate from all causes is about 40 ljercent higher than for nonsmokers. For those who smoke from IO to 19 cigarettes a day, it is about 70 percent higher than for nonsmokers: for those who smoke 20 to 39 n clay, 00 percent higher, and for those x-ho smoke 40 or more, it is 120 percent higher. Cigarette smokers who stopped smoking before enrolling in the seven studies hare a death rate about 40 percent higher than non- smokers? as against 70 percent higher for current cigarette smokers. Men who brpn smoking before age 90 hare a substnntiallr higher death rate than those who bepn after age 9~. Compared rlth non- smokers, the mortality risk of cprette smokers. after adjustments for differences in age. Increases w-lth duration of smoking (number of years), and is higller in those who stopped after age 55 than for those who stopped at an earlier age. In two stirdies which recorded the degree of inhalation, the mor- tnlitv ratio for :I given amount of smoking was greater for inhalers than` for noninhalers. The ratio of death rates of smokers to that of nonsmokers is highest at the earlier ages (40-N) represented in these studies, and declines with increasing age. Possible relationships of death rates to other forms of tobacco use were also investigated * * *. Th e death rates for men smoking less than -5 cigars n day are ahont the same as for nonsmokers. For men smoking more than 5 cigars daily, death rates are slightly higher. There is some indication that these higher death rates occur primarily 5 ward trend is reported in lung cancer death rates for the cntjre pcmp (smokers, es-smokers, and those ~110 never snroketl. co111l)ined) along wit11 n very sharp rctluc-tiou in cipwttc siiioking II\- the ph~sic~in11, is the Ibest n\ailnble PSillll~)lC of :I controlled wssatloll ~spriiiif~nt with retlurtioii of risks re511lt i11p fro111 rctliictioii of ~111oI;iuy. Tlie fintlii~gs of tliis report snpport the \.ic\v that cl)icleiliiolo~ic:il tl:\t:i sl~owll1g lower dent11 lY1tt5 :1111011g fol~lllel ~mokcrs tIl:lu amoiig cwitinaing smokcr5 cannot Iw tlisilriswl ai tliic to ~clwti\-e I)i:i~ niid that tlie lwne- fits of gi\-ing 111) ~111oliing haye l3rol~nl~ly IWCI~ iindci5t:ltctl. 5. Clgnrctte smokc~i~-: li:il-e lii~lirr ixtes of tliG;ll)ilit\- tll:ln n~~i~~ii~ok- PI'S, \~Ilt?tllt~~ llli?ilSlll~t~tl by tl;l\S lO?t fro111 \YOl'Ii :llllO;lf tll? Pllll)lO~Pd Iwpi~lntioii. 1)~ tlnys q&t ill'in Iwtl. 01' 11~ tile ino5t ,pnrixl i~i~wiire --dnp of "re.-;trictetl :Icti\-it?-" tliic to illncc~ 01' ilrjrir\-. T)at:l froiii the Sntionnl ITealtI1 Siir\-ey l)ro\~itlc :I I);ts(~ for e.+ti,rl:rti~y tliat in 1 yeal' iii tllr I'iiitctl StiltC'S aI1 :icl~litiorl;ll 77 1riilliol1 1Jr:ll~-tl:lj~i \\`el't' loqt flY)llI u-ark, an :~cltlition;il 8s nlillion 111:111-~l:i~.i \vert' 5lw11t ill i1r lwtl, alld ai1 :~dditinn:~l XX iirillioii i11;1n-t1;1~-~ of 1wti~ic~tcvl :1c`t ix-it F \wre cspe1+ e11cwl bccnn5c c*ig:1Wtte s;11lOkt'1.5 llil\.C' liigIlI(~r (lixllbilit,y 1':ltW tll:lll IlOll- smokers. For 11~11 :ye 4.i to (;4. 28 lwuw~t of tlrc tllwl,ility (lay- (IS- 1)erienced represent the excess associated with cignrctte duoking. In the 1067 Report tl1c follow in, m qliwt ion9 \wrv r~111I~li:1~izctl : 1. How much mortnlity and PSWSS diwl~ility :IW ;l+ociated with smoking 1 2. How nrwh of this earl- ljlort:xlit,y ant1 esce~~ tlisahility nollld not, linvc occ~n12ml if Iwople 11ntl Ilot t ;IliCll nl' ciprrt te :irroki1i,~? 3. ITOK mnch of tliis early nloi?:llity ant1 exws5 (li-:\I)ilit y ccailtl be nrertetl I)!- the crssntion or reduction of cipnrettc .SJllolii11~? 1. What are tlic l~ioi~ir~~linuisiiis R-herthy these eflwts take place and what arc tire crit icnl factors iii these niecl~nnisnis? The problem of ho\\- Iwst to me:1~11rc the relation-Ilip betn-een amok- ing and inortnlitv was presented by three menningfiil measiires of romparisoii : 1. Mortality Ratios: Obtained by cliriding the death rate for a clns- sification of smokers by tl1e death rate of a comparable group of nm- smokers * * * A nlortnlitv ratio has been considered to reflect the degree to wliictll :1 cIasxificGtioi1 \-:iri:\l)Ic itltwtifir-: or 111:iv acwmiit l'ol varintion~ in tlwtli rates. -1s s\ic*li. it is ;I Itiwqiiy of i.rl:iti'\-e risk v-liich indicates the importance of that, variable relative to uncontrolled vnri- ables-an iiicliwtor of jmf~~fifr/ /~;0/0~ji~ul .~;~q/1;fiuinw. 2. I)iffei~eiicw ii1 Jlort:llit\- IZntcs : Ol+t:\illctl I)\- 5liI)t rncting froiii the death r:ite for wroker5, the death rate of a c~~n~Ix113hle gronI) of nonsmokers * * *. Thi? measllre rcdlwts the atl~lwl l~rnl~xl~ilit~ of death in :I !-year Iwriotl for the ~111okcr c)\xar that for tile 11ow111oker. *Is such it is a 11iea~ilre of p1~1w~~~~7 7~~07th s+pi~~o7~~~i~. :I I~I~:II~S for the intli\~itln:1l to rStiil1atc tlie ntltlctl ri~iltic increace in death rates as age advances, with the incrtf~e being particularly- marked after age 45. The cleat 11 mtrs for coronary- lleart disease for n~en and u-omen continue to shoed il cwiisl)iruon tlitrernice. In 1966 it was Xl.6 for males ancl 22K:i for feirmle-; per lOO,OOO population. While seven~l studiw of wriow ns;pec.ts of the association between cnronar~ hc:lrt tlisense nlortality and viparette smoking have been rrportetl dilring tllr l,:tct year. tile. most significant studies of this association are cont:~inrvl ill the 1067 report. Tht se\-rral ne\v .stlltlici of variom aspects of the association between coro~ul:\- hr:rrt tlistme Inortality and cigarette making follow. Friedman ($6) reportetl a strong positive correlation betm-een per ml)ita c*igawttc sales ant1 ~wro~~:~~~ 1leilr-t disease death rates by states. The correlation is 0.76 when d:lta only from tllosc states with relatively acmratc information of cigarette consnnlption are. analyzed. Related factors sw11 :\s urb:\nization or softness of the local mxter supply do not explain this degree of association. 16 Other studies tleal with the excess deaths associated with smoking. Strobe13 et al. ( I.#) reported that among 3,479 Sviss physicians, over 30 percent of tlic exce.ss deaths occurring over a g-year period among siuokers n-as due to coronary heart disease. In contrast to the study above and c1nt.a from the T-nit& States in whirh nl~proximatelr one-half of the excess deaths associated with sinokiug are attrilmted to cardiovascular causes (I.$!?), preliminary data from Hirayama (6;;) show that the excess deaths in Japan associ- ated with s1ii0kiii.g were priniarily cxpl~ained 1)y cancer of various site,s. Only Ii! percent of tlie excess dentlis were associated with cardio- vascular causes. This prospective study of %.5,118 adults over the age of 40 encompassed ;I followup period of 15 months. -1dditional follow- up by TJirayanla slwuld yield useful tlnta with respect to smoking and excess mortality from cardiovascnlar diseases in this ,Japnnese popu- lation group, p:vtticularly with regard to the younger adults in the study. Hyams, et al. (67)) on the other hand, speculate that the apparent increase in the occurrence of coronary heart di+ase among +Japa~wse males, especially rmder tlie age of fifty, may be tlnc to a t,rend toward Westernization in both diet and smoking habits among younger Japanese men. Hammond (,54), in his prospective study of over 1 million men and women, showed a positive relationship between the duration of the smoking habit and coronary heart disease mortality. In the Framing- ham Heart Study (71)) no association wa.s found between the. duration of the smoking habit and the incidence of mortality from heart &tacks among men who were "heavy smokers" (more than one package of oigarettes per day). These diswepancies between t,he relationship of smoking to the incidence of t&al coronary heart disease and mortality from acute ooronary heart disease may be accounted for, in part, by the differences in population ,q~~~ps studied and by the possibility that duration of smoking may have a greater association with t,he fatal forms of coronary heart disease.. Knmrel, et al. (70)) in more recent. data from the Framingham study, indiwte that the fatal and more severe forms of coronary disease are more st,rongly associated with c.igarette smoking that the less sel-ere forms (figure 1). Coronary Heart Disease Morbidity * Rlnc.1~ of the morbidity da,ta reported during this past year resulted from retroqwti\-e studies of patients or cross-sectional studies (106, 107, 127, 134, 151). In these studies the findings revealed that there o Also may include mortality data in this presentation. 17 2.50 - *ing,na Pectorlr Myocardial Fatal Infarction CHO S"ddW Death 2.20 .g 1.50- E e 2 e s LOO- NOW kleavv Non- NW- Heavv NOW Smoker Smoker Smoker Smok;` Smoker Smoker Smoker S&k& OES. = number obsewed EXP. = number expected Heavy smoking indicates more than 1 pack per day FIGLYRE I-Morbidity ratios for .qxvified manifrstntions of coronary heart disease. nruonfi men aged X&SD years at entry into Frnminghnm Heart Stuclr, classified by smoking habit : 12 years' experience. ~OCRCE : Iiannel, et al. (70). were reht i\,ely more smokers among the groups with coronary heart tliSfXW, tll;\ll :~llIOllg tilt? ~Olllp\l'iSOll, 01' COlltrOl pOllpS. In n retrospective stitlltly of myocardial infarction patients in *Japan, I-Iy"ms, et :\I. (67) reported siinilnr findings, particularly among the mm untler age 50. DitFrrences measured 1)~ an exposure index com- bining intensity and duration of smoking showed the same trend, t.liougl1 the tl:lt:i were not statistically significant. Dorkcu (.M. .I/) reported on t\vo retrospective studies in Hamburg, Gerni:liiy : one: a study of female patients; the other, a study of male patients. IIe. cwnc-ludctl tllnt there is a strong association between w~oking :tu~l nivoc;lrdinl infarction in both males and females under the age of 45. In Dublin. JIulrnh~, et al. (106, 108, 109) studied groups of male and fem;lle coronary heart tlisense patients under age GO. He. found that a much greater portion of the patients, in comparison with a sample of the general population, smoked cigarettes. Also, the intensity (amount, multiplied by tlurntion) of smoking was RS much as 234 times gre.nter among the male patients and 3 times great,er among the fe.mnle coronary heart disease patients as contrasted with the males and females in the general population. In a study of 6'75 aviators, smoking histories taken in 1963 did not sllow a positive association in the prevalence of coronary heart disease with either amount, duration, or intensity of smoking. These findings are based on 38 cases (5.7 percent) of coronary heart disease of all forms among a very select population and are therefore subject to large sampling variations (9G). Moreover, since smokers may have an escwsi\-e mortality tllwing an acute nlyocartlial infarction. as mcn- tioned before, prel--alence rates are not as good a measure of the asso- ciation between snloking and coronary heart disease as are incidence r&es. Epstein (,?9), although finding no prevalence differences between smokers and nonsmokers in his Tecumseh Study, found an increased incidence in cigarette smokers of both fatal and nonfatal coronary heart disease. In a short prospecti\-e study of 14,000 Kern-egian men (12,000 with smoking histories), ?;atvig (113) did find an increased risk of inci- dence of first myocardial infarction or angina pwtoris among those men 50-59 years of age who smoked. Since the 1967 Report, t,he continuing prospective epidemiologic studies have comewhat clarified the differential relationship between smoking and each of the manifestation categories of coronary heart disease : angina, nonfatal myocardinl infarction, fatal myocardial infarction and sudden death. Data from the Framingham Heart Study (G9) revealed that "heavy" cigarette smoking, more than 20 cigarettes per day, is positively asso- ciated with uncomplicated angina in males but not in females (figure 2). Similar findings were report,ed by Weinblatt (155) in a study of male subjects in the Health Insurance Plan with t,he associations more pronounced among those men who smoked two or more packages of cigarettes per day. As can be seen, in table 2, the arithmetic differ- ences in rates between smokers and nonsmokers are greater for myo- cardial infarction than for angina; however, t.he risk ratios are similar. In a ret,rospective study, Heyden-Stucki et al. (61) found no asso- ciation of smoking with angina or other chest complaints. The inconsistencies in data on the association bet.ween smoking and the development of angina may be due in part to differences in methods used to diagnose and classify angina and to record smoking habits in these epidemiologic studies. Further standardization in this area may help to determine more accurately the relationship of smok- ing to angina. 19 2.00 - 1.61 l.!iO- ; > : l.OO- d L NCl"- Smoker OES. = "umber Observed EXP. = number expected NOW l-20 >20 Smoker CigaretteSmokers FIGURE 2-Angina pectoris morbidity ratios among persons aged 3&59 years at eotry into Framingham Heart Study, classified by sex and number of cigarettes smoked : 12 years' experience. SOURCE : Kannel, et al. (69). TABLE a.-Age-adjusted incidence rates per 1,000 males aged 35-64, and morbidity ratios, for speci$ed manifestations of coronary heart disease, by smoking category: Health. Insurance Plan Study 13 year observation data] Smoking catrgory Myocardial infarction Angina Incidence Morbidity Incidence , Morbidity rate ratio rate ratio I- --+---- Current nonsmokers ____ --~~---~ . .._.._ - 3. 27 1. 0 1. 37 1. 0 Current cigarette smokers_-_- - - ~. _ - ~. 7. 01 2. 1 2. 62 1. 9 Less than 2 packs_~~.--~~~-----~-~- 5. 05 1. 5 2. 08 1. 5 2 or more packs--~---_----_.- __._. 20. 80 6. 4 6. 64 4. 8 Source: Feinblatt, E. (f56). 20 In the Western Collaborative Study, Rosenman et al. reported !iipher rates of silent mvocardial infarct ions in ~~outlger men, 311d liigher rates of recurrent myocnrclial infarctions at all ages among those who smoked nwre than 25 cigarettes per day (13.3. 124). Friedemann, et al. (44) reported reinfarctions occurred more fre- Dorkeii (27) found in a series of 330 men of all ages, in Hamburg, who surriyed at least R and up to 6 years after their first. nlgocardial illfarction, that 172 (5" percent) had stopped smoking completely after the first, infarction. In contrast, of 8~ sul)jects who had died from a second myocardial infarction or sudden coronary death after learing the hospital, oi11y I&S (Z.9 ljercent ) had given up smoking completely (PSit elevations in blood l)ressitre, ln~l)itunl smokers teiid to 1ral-e lower blood pressures thall do 110111;11~01it?1'S. Hut, gi\-en the ljresence of high blootl pressure in an indi\-itliwl, smoking acts as an additional risk factor for the developnlent of coronary heart disease. Heyden-Stucki et al. (61) report that, among 500 workers in Slvitzer- land, smokers, particularly heavy smokers, have lower blood pressure as a group than do nonsmokers. Smokers also were found to have normal or subnormal weights in contrast to nonsmokers who had a greater mean weight ; thus, confounding the relationship between snroking and blood pressure level. Tibblin (144) in a cohort study of Scandinavian men born in 1913, found a lower mean blood pressure among smokers than among nonsmokers. As the population was clas- sified according to levels of blood pressure, a step-wise decrease in the prevalence of smoking was noted as the level of blood pressure in- 21 TABLE 3.-Mean age and man systolic and diastolic blood pressure, by sjnoL+irLy category: Los Angeles Heart Study, 1962 Current cigarette smoking status Blood pressure (mm. of Hg.) Number of, Years of Systolic Diastolic subjects age -1 -,-,- I Smokers.~~~~~~.~~-------~~~---------- 407 54 133. 6 82. 5 Nonsmokers--.~~-------~~.------~-~~~- 728 57 ~ 137. 0 83. 9 I SOURCE: Clark, V. A. (2s). creased. A similar trend for both systolic and diastolic pressures was also reported by Clark, et al. (23) as shown in table 3. In the study of 675 arintors (96) smoking intensity, although not found to be associated significantly with systolic or diastolic blood pressures, was positively associated with pulse pressure. Reid, et al. (122) in a comparative study of workers in Great Britain and the United States noted lower diastolic blood pressures among smokers than among nonsmokers in both groups; adjustment for Tveight variations reduced this difference appreciably. Rfulcnhp (207)) in a retrospective study of 100 women coronary heart disease patients under 60 years of age, reported that ,50 to 60 percent had diastolic. hypertension (>nO mm. HF.). Hypertension and ciga- rette smoking, together or separately, were present in over 80 percent of these patients. In the major prospective studies, when both smoking and hyper- tension were present, an interactive increase in the risk of developing coronary heart disease \yas noted. When to these two risk factors elevated cholesterol levels were added, the risk of developing coronary heart disease was further increased (figures 3 and 4). High A`Temm~ Cholesterol ad Relcrted Diet Certain of the retrospective and cross-sectional studies (6'9, 151) have, in general, demonstrated higher cholesterol levels in smokers than in nonsmokers. Pincherlc, et al. (119) and T,ane. et al. (9G) report similar finding?. Ai stud7 by Heyden-Stucki (GI) of ,500 Swiss workers found a similar trend but the differences between smokers and non- smokers with respect to cholesterol levels and other lipids were not statistically significant. A recent report (A'/;) describes some of the variability of interrela- tionships among smoking, blood pressure and cholesterol levels in different population groups througllout the world. It concludes that though nonsmokers tend to be heavier and have higher blood pressure levels than cigarette smokers, heavy smokers tend to be in the top 22 150- positive weight W.C.H. cd" any 2 any2 any 2 any 3 any 3 OnI" only On," only or all ""I" O"lY OnI" or 3cml" all 4 CHD 46 1 0 2 8 4 6 12 17 23 N 1,329 94 161 227 260 159 204 484 222 264 6 42 Relataonship between status with respect to four coronary risk factors thyper- cholesterolemia, hypeRen~hm. overweight. and cigareffe smokingt as evaluated on original examination and incidence of cltn~al coronary heart disease in men or,ginally age 40-59. free of definite CHD. and followed subsequently wtfhout systematic interventgon. Peoples Gas Light and Coke Company study. 1958-1962. W is overweight, le. a ratio of observed weight to desirable weight of 1.15 or greater, C is hypercholesterolemla. ie, Swum cholesferol level of 250 mg/lOO ml or greater. H is hypertensmn. ie. a diastolic blood pressure of 90 mm. Hg or greater: S is smokmg of ten or more cigarettes per day.' FIGURE 3-1ncic:enre of coronary heart disease among men aged 40-59 years at entry iuto I'ctiol~les Gas Light and Coke Company Study. classified as to pres- ence of specified risk factors: l!Xbl962. SOURCE: Stamler, et al. (1%). cleciles for blood pressure nnd rrlntiye x-eight. Cholesterol-smoking relationships described in these studies do not show a consistent pattern. Tn a controlled dietary intervention study of postjnfarction patients Leren (W) fount1 that smokiiq Ilabits did not iilflilence the serum cl~olesterol level or tllr cwronary llci1rt clisease relapse rate in the colltrol group. i~lllollg the stlltl) gronl) of clictcrs there was il suggestion, :~lthough not statid ically >igl~ilicxlit at tile 0.05 level. that smokers had a higher coronary heart disease relapse rate than nonsmokers. Physicd rnnctic4y The independent and combined effects of cigarette smoking and physical :ic.tiCty, as dr~~ribetl in the 1!)67 rrlwrt. vontillne to be demonstrated :IS more data are :Ic~,,l~llllllntetl. Tile apparent l)rotcctire effect of plrysical :ic*tiritJ appears to be more pronoiinwd with regard to myocarclial infarvtioll tlian :lllgiu:l [t;Ible 4, (255)]. Differences in methods of assessment of liistoq, of l~ll~sical acti\-ity ill cape versus 315-1310-66-3 23 6.00 - 1 .w - 0.00 Any one Any two Any three OBS = number observed EXP = number ex~ecfed Number of abnormalities FIGIXE 4-MFocardial infarction morbidity ratios among men aged 30-59 years at entry into Framingham Heart Study, classified according to presence of selected riak factors : 12 gears experience (Risk factors are: cholesterol lerei over 250 m&100 ml., systolic blood pressure over 160 mm. Hg., smoking over 1 pack of cigarettes per day). SOURCE: Kannel, et al. (70) control groups may account for some differences in the incidence rates noted. Hlackl~urn, et al. (10) folmd no relationship of smoking to the prev- ale,nce of postexercise KCG c.hanges in a study of 10,260 men age 10 to 59 years. TTo\vever. there were only .51!1 (.i.l percent) subjects with a, "positi\7" TSCG response. Sociologirn?. Pqvh~olqicn7 and Pwsonalify Variables Tn-o studies ($5. R/t) demonstrated an inverse relationship between t,he frequency of coronary heart disease and the educational level of the subjects. In the Bell Telephone System (Cd), those men without a college education had higher coronary heart disease rates than those with a college education. Also, those not at.tending college tended to smoke more. In a study of factors related to coronary heart disease among Cleve- land attorneys ($5)) the quality of the law schools attended by the sub- 24 3. 27 1. 0 6. 33 1. 9 2. 14 1. 6 3. 07 0. 9 1. 67 1. 2 3. 01 0. 9 1. 32 1. 0 7. 61 `2. 3 4. 71 1. 4 3. S.i 1. 2 11.27 3. 3 24. 09 i. 4 2. 0.i 1. .j 2. 37 1. 7 1. 9.5 1. 4 4.97 3. 6 5. 09 3. 7 12. "0 3. 9 jec.t were ranked independently by a law school professor. I,a\~-~ers attending s~~l~ools in tile "lliglrest law ~:cllc:~ol quality growl)" Ilad lower rates of coronary heart disease than those attending schools in the **lower law school qlwlity group." *use, t11ow ill tile 1attrr group had started snloking at an earlier age. Since additional differences were noted for other risk factors, smoking alone may not be responsible for tllr total differences in these rates. In both stnclies, it uxs liyl)othe- sized that with respect to susceptibility to derelopmellt of coronary heart disease, behavior patterns and attitucles estal~lishetl l)rior to professional traillillg and prior to stresses rcsnlting from job mobility and job tension, were nlore significant than the later stresses x-hich resulted from their present jobs. Rec,ent clata from the Western Collaboratire Group Stll(ly (IX) appear to show that among met1 X-49 Jears of age. cigarette smoking ~-as associated with several coronary heart, disease risk factors (table 5). Though these findings may be statistically significant, the differ- ences between smokers ancl nonsmokers I\-ere small. 25