Tobacco Documents Online

William B. Kannel and William P. CosteUl There is.now ample evidence to show that nicotine absorbed from inhaled tobacco has acute„transient effects on the circulation which could explain many of the observed epidemiologic features of the reiation of cigarette smoking to the development of, cardiovascular disease. Its actions are eompatabte with a transient, non-cumulative and reversible triggering effect operati ve in persons with an already compromised coronary, circulation. The high carboxyhemoglobin values generated by cigarette smokers also fitsthis pathogeneticconceptuali¢ation by causingiiurther impairment of oxy,geniudlization by ischemic tissuessupplied by critically narrowed . 76 1250 369 SIGNIFICANCE OF' NICOTINE, CARBON MONOXIDE AND1 OTHER' vessels and catechol stimulated by nicotine. In additioni efifects of nicotine on platelet adhesiveness and other clotting factors coutd, impair flow in the micro- vasculature or promote thrombosis,in near occludedl arteries. Evidence which in* criminates ciprettes in the process of atherogenesis isiess substantial and does not' explain as well'the epidemiological relationships of the cigarette habit to the oc- currence of cardiovascular disease. Ailthough it is quite likely that severe exposuree to carbon monoxide promotes atherosclerosis in the heavy smoker, it is more like- ly that the acute precipitating effects of nicotine and carbon monoxide are re- sponsible for the excess risk in the cardiovasculardisease-prone smoker. Giving up, smoking has been shown to reduce risk,by about,half so there is much to be gained by abolishing the use of cigarettes or in reducing their nicotine and carbon mon- oxide content. There should no, longer be any lingering doubt that cigarette smoking is as- sociated with an, excess rate of occurrence of cardiovascular morbidity and mor- tality (11-3). Because the relationship is independent of adl the known contribu- tors to cardiovascular disease and since giving up the habit is associated with a substantial reduction in risk,:a causal relationship seems likely (4-7),. A number of'pathogenetic mechanisms have been postulated toexplain how cigarettesmok- ing induces cardiovascular illness (4, 8, 9)L The action of cigarette-induced acute and chronic nicotine and carbon monoxide poisoning has been most prominent- ly incriminated by investigators (2, 8, 9;10),. Mechanisms postulated to explainihow cigarettes promote cardiovascular dis- ease must be consistent' withi the known facts aboutthe relation of'the cigareite smoking habit to the development of the atherosclerotic cardiovascular diseases. These facts may be illustrated fromiFrarninghartt data and summarized as follows: The risk is dose related to the number of cigarettes smoked per day (Fig. 1) . Althoughi there is some evidence to suggest otherwise (6, 1'2) the risk may not be related to the duration of the cigarette habit (Fig. 1!). ~~SIV10'~KE~~CIO'MiPO'NENTS IN THE DEYELOP~MiE'NT'O'~F~ CARDIOVASCULAR DISEASE I

KANNEL A(WDCASTE'Llll. ~. ..ra..,a ue i... .. ~..1 ~. yRrAY~~~ m~~~~ ~ I ,,..,........,... .e .,..a ~~ ~~f~Yll1{yIAV. ~IN~. ItN ® f IG. 1. Risk of wronary, attacks according to number of cigarettes smoked pen day vs. dura- tfon of ci=arette habit-Fiunlngham Study. ._.._ ,_ - The risk of coronary heart disease in the ex+smoker may be as low or lower than that of'ttiose:who never smoked (4). Giving,up the habit is associated with a, prompt reversal to lower risk, although some contend that the benefits acuue more gradually (9, 11,,12)',. In any event, those who quit the habit have been found to have only haif'ttie risk of those who continued to smoke (Fig. 2). The :o Is I I sWoMMa. w.ar RtK!!OM Oo0*. snwow trwoe O.HO o.Po. o.ar, o: m. OJ P6' 0229 Tr~IU[' l.lt :..T 107 .T I ; 1 FIG. 2. Incidenee of'eotonary attacks among ciyaretta smokers accordfn; to subsequent clt- arette habit-h1ee~45-74 at exam: Framingham Study-1S yeariollow-ip.

SMOKE COMPONENTS AND CVD 371 benefits of' quitting wane with advancing age and no longer accrue beyond age 65' (7);. The contribution of the cigarette habit to coronary risk in women is feeble (Table I) and in both sexes diminishes with advancing, age (Table I I'). In men ciga- rettes are significantly related to all cardiovascular endpointsexcept possibly con- gestive heart failure. For women in the Framingham cohort only in brain infarc- tion is there no apparent association (Table IIl). TABLE I Average Annual Incidence of Coronary Attacks Per 10,000 Persons at Risk According,to Cigarette Habit. Men and Women 45-74 Framingham Study: 18 Year Follow-Up „ Men Rate ot Coronary Attackt No.ol'Ggorettes Per Dny Person- Yrs. At Rlah No.oLNtw Coronary Attakcs Crude Smoothed Actual Age-Adjusted Age-Adjusted None 11552 85 ' 74 70.6 67.0 Uhder 20 3938 40 1102 90.7 98.3 20 S3'24 71 133 116.3 137:1 pver 20: 5096 59 Women 116 148.8 125.6 None 22802 70 311 27.4 27.6 Under 20 6986 22 31' 3IIS 35.4 20 3754 10 27 38'.6 33:5 Over 20, 1610 5' 31 452 29.3 T-Vaiue-For age:adJusted rate: Men = 4.77 Women = 1.50 Source: Framingham Monographi#30,. TABLE 11 Cae ucient forthe Regression of Incidencrof'Coronary Attacks"'on Number of Cigarettes Smoked Daily According,to Age in Each Sex. Meniand Women 45-74 Framingham Study: 18 Year Follow-Up Men Women 45-54' 33-64 65-74 45-54 SS-64 65-74 Regressfbn Coetflcients. .342' .227 .039 .313 .123 -.142 T-Values 3.73 295 .0-27 1.54 0.76 -0.47 N:umberofEvents• 92 121 45 22 SO 36 'Goronary attacks: Manifestations of coronary heart disease other than angina pectoris. Source: Framingtiam Monograph #30.

,. 372 KANNEL AND CASTELia TABLE' III Regression of Incidence of Major Cardiovascular Diseases onPlumber of' Cigarettes SmokedlDaily: Men andlWomen 45-74 Framingham Study: 18 Year Follow-Up ' Msn Regression Coronary Brain Intermittenr' Congestive Totol Coefficient Attacks Infarction Claudlcation Failure C: V' Olteast BivarJate' .255 .318 .375' .080' .209 Multivariate*" .300 .373 .410 .148' .237 Multivariate ThValues 5.48 3.011 4.67' 1.61 '6,27 Women. Bivariate• .173 -.096 374I .253' .063 Niultivariate" .227, -.0'1'3 .5'07 .314' .134 Multivariate T-ValUes 193 -0A0 3.431 2.46 191 'Bivariate: Number of cigarettes and age. •*Muitivariate: Number of cigarettes andagepius-systoiic blood pressure,serum cholesterol, glucose tolerance, ECG-LVH. - Source: Framinghami Monograph `TM`301 The impact of the cigarette habit is stronger, for some cardiovascular end- points (9) (such as occlusive peripheral arterial'disease) than others (sucfi as con- gestive heart failure) (Tables II,,III): In coronary heart disease:the relationship iss strong, for sudden death, moderate for myocardial infarction and non-existent for angina pectoris (Table IV'): • , TABLE IV Incidence of Specified'Clinical Manifestations of Coronary Heart Disease According,to Cigarette Habit at Each Biennial Exam. Men 4'5-64 Framingham Study: 18'Year Foll'ow-Up N'o.,of Cigarettes Averoge Annual,Lncidemce Per 1G1,000-Age-AdJusted Per Day Each Coronary Myocardial Coronary Sudden Angina Biennial Exam. Attocks•' InfactJon'•' InsuffJciencyt Dieath• PecrorJt* None 50 32 8 ~ 8 43 Under 20 83' 53 10 25 35 20 Over 20 118 ' 81' 7 20 35 1106 68 14 22 51 1-4 Q' tNotstatisticaiiy significant at P=<.05. •P- C.05 '~T ~1 **P - <.01 . ~ Source: Framingham Monograph #'30. ~ ~ V.1 . The relationship of the cigarette habitto cardiovascular disease is independ- ent of the other majwr contributors to risk and it is most ominous in those pre- disposed by other risk factors (Table Illly Fig. 3);

SMOKE COMPONENTS AND CVD 373' 1 ~. ~ .. W Y >,0 t: 4 O O c ~ 1s9 sx M 0 70.e I" 333 t • FIG. 3. Probability of, developint, cardiorascular ditsase in,a'. years--clPrette smokers vs. nonsmokers accord(ng to Ibvel of'other risk factors-40 year oldlmen: Framineham Stud!y- 1E year, followwp. SYlfOtlC Iu000: rllcseNRl0g .1!I! I.ssWrol4OLtaRCna' rs1M. 3311' . QilUWtaNfOl[wIK9 0' O. 0 tc0-lx" 0 0 0 SO{ACG YOMO0IWM M0. a~' T0 so so. 40 30 20 10 There is littVe$, if any, increased risk associated with cigar and!pipe smoking,, presumably because the tobacco is not inhaied (4). While there is a strong association in most' affluent' societies, there is a, lack of'demonstrabie effect of cigaretrtes in some low coronary incidence populations (Fig. 4). . = M. • E l / ' ~. i ~. ._ FIG. 4. Probability of'CHO occurrence In two years according to amount of cigarette unok- ing-Framin2ham„Monolulu and'Puert'o Rlco-Men 45-65.

374 KANNEL A~MD CASTELLI The postulated pathogenetic mechanisms must' be consistent with these facts. They must therefore. explain what appears to be an independent, transient non- cumulhtive,,reversible triggering effect that appears to operate impersons with ani already compromised arterial circulation in a fashion tharfavors sudden occllrsionn of the circulation to the head, heart or limbs or promotes lethal cardiac dysrhyth- mias (4; 8, 9) rather than one which primarily promotes the underlying athero- scierotic process. TOXIC PROPERTIES OF' THE CIGARETTE The awesome number of'toxic substances have been isolated from the ciga+ rette (13; 14). Among the gaseous and'particulate substances released when ciga- rettes are burned and inhaled are some which have been, demonstratedlto at least acutelif affect card'iovascular, metabolism, function, and physiology (8,9,,13;14). Foremost' amongst those identified are nicotine and carbon:monoxide (!8, 9, 1i0). Apparently of lesser importance are heavy metals,,trace metals„nitrogen dioxide and substances whichi provoke autoantibody and non-specific inflammatory re- sponses (15). Cigarette smoke contains in the mainstream of the gaseous phase- not only CO; but 2'50 p:p.m. ofnitrogen dioxide (or dinitrogen tetraoxide) sub+ stances of concern, to environmental' protection agencies. The safe level of this gas (5' p.p.m.) is easily exceededby many cigarette smokers. Although this noxi- ous gas has been incriminated in lungdisease,,no connection with cardiovascular disease has been demonstrated. Pernicious adverse mechanisms affecting, the cardiovascular system, which can be invoked from the known properties of inhaled cigarette smoke include: hemostatic effects, cardiodynamic influences, atherogenic results, and vasculo- toxic or inflammatory responses (8, 9)',. These are: among the known cardiovas- cular effects of nicotine and carbon monoxide which must brconsidered. Nicotine From 0:3-3.0 mg, of nicotine can be recovered' from smoke taken into the mouth from a cigarette. The amount absorbed wilUvary from 5% in thosewho do not inhale to 10096lin!the deep inhaler. A heavy smoker who inhales is subjecting himself to the influence of 50-100 mg of nicotine daily ('8).The principal effecrof this absorbed nicotine appears to be to stiinulatrcatecholamine production and release of noradrenaline from local stores (8, 9),. This prodwces sympathetic over- activity provoking cardiodynamic effects manifeste&by a rise in heart rate„blood pressure, cardiac contractile force, cardiac output and; as a result, rnyocardial oxygen demand ('8; 9). Inihealthy persons this is compensated for by a compensa- tory increase in coronary blood flow„but in those with atherosclerotic stenosed coronary arteries this may not occur and ischemiaresul!ts (I8). Indeed,aftera myo- caidial' infarction, smoking may cause cardiac output and stroke volume to fall' ('1 G)'. The high myocardial oxygen consumption of the cigarette smoker may-not be entirely explained on the basis of the 'nicotinc-catecholamine induced enhance- ment of inechanical' activity of_' the heart and about half; may result from the metabolic stimulation oPthe high concentrations of free fatry acid's generated!(17).

SMOKE COMPONENTS AND CVD 375. In contrast tioi these eardiodynatnic effects, few atherogenic effects of nico- tine have beeni convincingly demonstrated. Even when adfninistered'in amounts relatively. higher thanithe nicotine uptake of a human smoker, no atherogenic ef-' fects have been shown inianimals ('18)', Some,, but no't' all studies, have foundia il rise in serum cholesterol in cigarette smokers (8), On the other hand, an increas+e li in circulating free fatty acids has been consistently demonstratedlwhich may later result in increased triglyceride, or possibly, cholesterol levels ('1'9)'. Such athero+ ~ genic effects as there are seem minor, reversible and the findings arenoz consis- ~i tent. More importantly, either as a consequence of the rise in free fatty acid or ' i as a direct effect of noradrenalin, myocardral irritability is increased which may I, predispose to sudden death (8, 9, 20). ,. •' IIIII Nicooine-inducedl release of catecholamines also has hemostatic effects and can increase platelet stickiness and aggregation thereby accelerating thrombus I formation in damaged vessels, possibly contributing to the development of athero- mata andl enhancing thromboembolic sequelae in personswith advanced athero- sclerosis (9, 2'1'),. The~smoking of a cigarette has been shown to increase the plate- (et's' response to a standard aggregating stimulus. This phenomenon,, which is i specifically related to the inhalation of'tobacco smoke is not' related to carbon monoxide or to the particulate : matter in the tobacco smoke. Nor does it' appear dependent on the rise in plasma free-fatty acid provoked by cigarette smoking (22'): Evidence to date suggest that thys platelet response is a direct consequence of either the niiaotine itself or the catecholamine response: it elieits (22): This smoking-induced potentiatuon of platelet aggregation could' go a long, way to- wards explaining premature occlusions of stenotic arteries in cigarette smokers. Other hemostatic effects claimed in about 100 papers since 1963 include: de- creased fibrinolysis, decreased clotting time, change in the rateoPinitiaV clot for- mati©n, maximum clor tensile strength andiclo't'retraction (23)s In the brain,,nicotine as well as smoking,inereases cerebraliblbod flmw while at the same time decreasing the arteriovenous oxygen difference, leaving the cere- bral metabolic oxygen rate unchanged (24). - . i Cor6o Nl de l n onnxr Cigarette smoke, in addition to nicotine, also contains from 2.7-6.0% carbone monoxide which when inhaled exposes the smoker to 400-500 p.p.m. of CO, a, level eight times that permitted ini industry (I8). The hazards of CO have been recognized for many years, but only recently connected with the cigarette habitt which now appears to, be the chief cause of chronic CO intoxication. The CO levels encountered in city traffic range from 20-200 p.p.m.exposingnpn-smokers to a T,3'S o carboxyhemoglobin content. Such an exposure is trivial compared to that self-induced by smoking. Because of its greater affinity for hemoglobin than ozygenj CO converts from 3-1596 of smokers' hemoglobin to carboutyhernoglobin. Carboxyhemoglobin not only decreases the-0i capacity of the blood but also al!ters itsoaryhemogliDbin dis- sociation curve so that oxygen is released'more reluctantly to the tissue because of tighter binding,(25, 26). Carbon monoxide may aVso interfere with tissye oxy- gen metabolism because of its high affinity to myoglobin (27,28) which normally t

376 KANNEL AND CASTELLI takes part in the transfer of oxY$en from hemoglobin to tissue mit+achondria and'o also plays a role in storageof oxygen. Through hypoxia CO acts to increase the permeability of'the~ end©theliumi allowing deposition of cholesterol (29). No sig- nificant change! in arterial lipid synthesis has been observed (30),. Thus by mecha- nisms different than those inditced by niicotine„CO may also adversely affect the cardiovascular system. The vessel' damaging effecu of carbon monoxide and hy poxia which allow deposition of cholesterol in the vascular in!tima has beernex- perimentallydemonstrated in animals (31I„32). All the atherogenic changes'attributed to carbon monoxide-increased endo- theliali permeability, accumulation of lipid in the vessel wall and anatomical changes-can be produced by hypoxia alone. Severall investigators have found increased'hematocrits in bothi experimental animals and man as a result of cigarette smoking (33,, 34', 35). This has been shown to result from an increased!eryt'hrocyte mass.(3~6) and may at times reach polycythemic proportions (37). This smoker's erythrocytosis is evidence that. chronic carbon monoxide exposure results in significant tissue hypoxia. In the Framingharn cohort, within the normal range of hemoglobin, valltes,,risk of cere- bral infarction was found to be proportional to blood hemoglobin concentratiom in both sexes ('Fig; 5),. Higherfiemoglobin valuerwere also foundito beassociated withi higher, diastolic blood pressures (38): Eff.ects of carbon monoxide on the brain, other than lethargy and obtundedlsensorium are notwell described. ® ! IN i FIG. S. Risk of cerebral lnfarctlon (9!6 year follow-up) according to antecedent hemoglobin and blood pressure status-Men andlwomen3l0-6II at entry; Framingham Study. More important than its atherogenic effects, the carbdn monoxide inhaled with tobacco smoke reduces the amount of oxygen available to vital.organs such as the heart at a time when its work and demand for oxygen has been stimulatedd by nicotine (25): In, normal persons carbon monoxide will increase coronary blood flow as nicotine does. On the other hand, like nicotine it may produce myocardial hypoxial in the presence of severe coronary atheromatosis: This has been demonstrated clinically in humans by Aronow who showed ttiat'lessezercise is needetd to induce angina after smoking (39). This effect is'very likely a, conse- quence of'CO, induced impaired oxygen utilization in the presence of'critically l

SMOKE COMPONENTS'ANiD CVD 377 narrowed vessels which cannot dilate inia heart'which at the same time has beenn stirmuiated by nicotine. V1laldletol have provided some epiderniologicai data directly connecting,car- boxyhemoglbbin levels: and the frequency of coronary heart disease (40)L It is hard to adduce! from this whether the risk is actually related to the CO per se or to the daily dose of other substances simultaneously inhaledlwithtobaccosmoke. Hmwever, experimental studies show that smoking•induced elevation of car- boxyhemoglobin may adversely affecrcardiac work performance, induce ischemic ECG changes and dysrhythmias in persons with clinical or subclinical coronary disease: The carboxyhemoglobin value may also be a more objective means of determining risk in relation to smoking,thana personal history of cigarette use. A maximum of'about 15% carboxyhemoglobin has been recorded in heavy cigarette smokers. Most pipe and cigar smokers who.presumably do not inhale, achieve a 1-2% carboxyhemoglobin level. However, converted cigarette smokers who take up cigars often continue to inhale and may indu+ae carboxyhemoglobin: levels of: 20% or more, a level capable of prodtacing symptomatic acute and chronic effects (41). PATHOGENETIC IMPLICATIONS Thus, there is ample experimental evidence w show that' the nicotine ab• sorbed from inhaled tobacco has acute„transient effects on the circulation which could explain many of the observed epidemiological features of'the relation of cigarettes to cardiovascular disease. Its actions fit fairiy wellff with a transient, non-cumuliitive, reversible triggering effect operative in persons with an already compromised coronary circulation. The observation of high carboxyhemoglobin values in cigarette smokers also fits this pathogenetic.conceptualizatipn since this could further impair oxygen utilization by ischemic tissues supplied by critically narrowed'vessels and catechol,stimulated by nicotine. Likewise,,effects ofrico- tine on platelet' adhesiveness andi other clotting parameters could'inflWence flow in rnicrovascul~ture or promote thrombosis in near occluded'vessels precipitating vascular catastrophes. Evidence which incriininates cigarettes in the process of atherogenesis (lin- volviing nicotine and carbon monoxide)~ seems to, these reviewers Iess substantiaV and does not explain as well the known relationship of'the cigarette habit to the occurrence of cardiovascular disease. It is qµite:likely that severe exposure to car- bon monoxide (in the very heavy cigarettesrnokeror inhaling cigar smoke) pro- motes at'herosclerosis, but it is more likely that the triggering or precipitating ef- fects of nicotine and carbon monoxide are responsible for the bulk of'the excess risk in the smoker. It is more reasonable at' this point to interpret the cigarettee habi't as playing,a contributory rather than a primary roie in cardiovascular'dise'ase. This is not the case in emphysema and lung cancer.

. N 378 KANNEL AND CASTELLI' THE'RAPEWT'IC~~ AND PREVENTIVE IMPLICATIONS Aside from abstention, soJutions to the cigarette health problem may be sought in producing, a safer cigarette-i.e., one with~less nicotine and tar and one! which generates less carbon monoxide. The possible beneficial effect of a less harmful cigarette or of giving up the! habit is difficuh to estimate precisely in the absence of data from a eqntrolled clinical trial. If one examines the attributable risk for cigarettes (i.e., incidence of C-V' disease in smokers minus non+smokers divided by incidence in smokers) we would attribute as much as 33% of cardla .vascular cases to the cigarette habit (Table V). If the effect were cornpletely in- dependknt (as it seems to be) and ?otally and permanently reversible (as it may be) one would expect a 33% reduction in~cardiovascular mortality from cessation of smoking. While there is no secure basis for making such predictions data, from Framingham seem to indicate that the estimated benefitsmay not be unrealistie. TABLE V Attributable Risk for Cigarette Smoking,in.Cardiovascular Disease. Men 45-74'. F'ramingham Study: 18 Year Fo(lowJJp. 1 t CardlowscuJar Dlfcase Cigonttts Awerogr Annual'Jncidencr Per 10,000 Per Day' 45-54 SS:-64 6d+74' All Aqa (yl djusted) None 79 203' 293 150 20 172 3'37' 3011 223 Difference 93 134 . E 73 A'tuibutable Risk (%) S4' 40' 3 33 Coronary Attacks Gqoretres Averoge Atinual'lncidence Per 10,000 Per Day 45-54 55-64' 65-74 AJI Agu (Adjusted) None 33 90 ": 14S'1 71'. 20 109 169 163' 116 D(fference 76 79 1d' 4S' Attributable Risk (96) 70 47 11 39 For coronary attacks one would estimate from the attributable risk (Table V) a 399o reduction should occur and this is in fact close to what was observed (Fig. 2). Thus; this is not a triivial'. issue and there is something substantialito be gained in identifying,the harmful!ingredients in cigarettes and removing them or abolishing cigarette use. The problem of devising,a safer cigaretteisapparently'tecFinologicalVyachiew able without great, difficulty. Getting the public to use them may be a more dif-flicult matter since the harmful ingredients may weil'be the ones that give the habitual smoker his satisfaction. Turner eto/ fiound thatonichanging to-very low I • . t