Tobacco Documents Online

Testimony of STANTON A. GLANTZ, Phl) Pro~.sor of Me~cine Member) Cardio~mcular Res~mv.h Imtitut~ Member, ImtiW_t~ of Health Policy Studies University of California San Francisco, CA 94143-0124 regarding Environmental Tobacco Smok~ prepared for Occupational Safety and Health Administration August 11, 1994 My name is Stanton A. Glantz. I hold a Ph.D. in Applied Mechanics and Engineering Economics fi'om Stanford University, where I wrote my dissertation on the mechanics of cardiac muscle. I am now a Professor of Medicine at the University of California, San Francisco in the Division of Cardiology and a member of the Cardiovascular Research Institute and the Institute for Health Policy Studies, as well as th~ Graduate Group in Biostatistics. I also serve as an associate editor of the Journal of the American College of Cardiolo~, the largest clinical cardiology academic journal, and serve as a member of the California State Scientific Review Panel on To0dc Air Contaminants, a body somewhat analogous to the federal EPA'¢ Sdence Advisory Board. I am here today to discuss the reasons that passive smoking causes heart disease and also to summarize work done on the economic impacts of 100% smoke-f~ec restaurants. In recent ycars,-most of the public discussion on the health effects .of ~ smoking has dealt with lung cancer, probably because of the ~ ~ by the tobacco industry regaling the 1992 Eavh-onmental Protection Agency risk assessment of 1 T102322.204

envi,.-onmenml tobacco smol~.c (ETS) and lung cancer (1). In fact, hcs..-t disea~ is acma~ a much more important endpoint oi~ passiv~ smoking than lung cancer (2-7). 'Whereas E'rs cauls 3,000-5,000 lung cancer deaths annually, it causes 30,000-60,000 he~_n disease deaths annually. The increase in individual risk of heart disease death due to ETS exl3asure is similar to lung cancer; the much larger number of attributable deaths for heart disease occurs because it is much more common than lung cancer. The American Heart Association has concluded that ETS is a major risk factor fi3r hear~ disease in both adults (8) and children (9). The biological and sdentific situation surrounding heart d.i.eease is qualitatively different from that of cancer, including lung cancer. Whereas cancers take a long time to develop, making it essentially impossible to demonstrate an effect of e~posure to a specific carcinogen in an individual, it is possible to demonstrate effects of envimmental tobacco smol¢~ (El'S) and its constituents on the cardiovascular function of specific individuals with relatively short term e.A'posums. This information complements population-based epidemiological studies demonstrating that passive smoking increases the risk of death or morbidity from heart disease. The reason for this is that, while some of the effects of smoking on the heart, such as the induction of athems~:lerosis (narrowing of the coronary arteries because of build up of fat deposits), occur over a long period of time, many of th~ other effects, such as alterations in blood chemistry, effects on blood vessels and the heart muscle itseE, and the ability of the heart to respond to the stress of ischemia (reduced oxygen delivery) occur immediately upon exposure to stooled. Th~se effects have b~en observed both in people and in laboratory stodies. The ability to induce negative ch~ in the cardiovas~ar system by short-term e~xmum to secondhand smoke is impormm from T102322205

heart disease. From epidemiological studies a very consistent picture of an devotion of about 30% in risk (for both death and non-fatal coronary events) has emerged, even after controlling for other known cardiovasaflar disease risk factors (7). These population studies are, in fact, more consistent in thei~ findings than even the studies demonstrating that passive smoking causes lung cancer. Many of these studies also show a dose-response relationship; people who are exposed to higher doses of secondhand smoke suffer more death and ~ab~l/ty from heart disease. The consistency of these studies, par~culady a~ter controlling for other risk factors for heart disease, together with the fact that they are done in many different countries with dif~ring diets and lifestyle factors, the presence of the dose-response relationship in many of the studies (10), and the fact that active smoking causes heart disease in smokers, would be enough to conclude that passive smoking causes heart disease in nonsmokers and to proceed with a quantitative risk assessment. An Overview of the Effect of Secondhand Smoke on the Cardiovascular System The cardicr~ascular system transports c~gen fl'om the lungs to the organs of the body and transports waste products of cellular activities so they can be removed by the kidneys, liver and lungs (11). In contrast to other muscles in the body, which only have to work some of the time, the heart muscle contracts continuously, around once every second, throughout a person~s entire [fie. The fact that the heart muscle has to operate continuously makes it d~!ferent in speci~zed ways from oth~ muscles. Unlike other muscles, the heart muscle has a con "tmuous demand f~ u~ygen, whk:h the heart muscle ce~ (like all cells) turn into 3 T102322206

c~a .tracdoa and other ~ process~ Coatrol of t~ heart mmd~ is ~ ~e~ adj~. It d~s ~ ~u# a ~mb~fion of ~o~ ~at m b~t ~to ~c s~ of ~c ~cr a lo~er period of ~e, o~cr ~ ~ ~ ~ m~me m ~ dem~ds on ~e ~~ ~ For ~ple, ~ ~ M~ hem e~n~y at ~a our bodies ~d adapt by ~~ h~t~t Im~ ~e ~on of bl~ ~pied ~ ~ese ~ ~mple~ sm~g up or m~ to Den~r, ~ but ~ ~ ~t ~e caMi~l~ ~tem adap~ bo~ ~ ~e ~o~-~d lo~-te~ to ~g~ ~ en~,n~ conditiom. Indee~ ~, c~i~~ ~mm ~ so ~lomly ~apfi~ ~m ff ~n~ues to function enou# to ~ ~ e~n ~cr ~c hem i~ ~ scfio~ly d~g~ ~ a h~ attack or if ~c en~mcnt ~ wMch ~c h~m opcrmcs ~ s,fiomly comp~cd ~u# cithcr cn~romcn~ t~ or o~cr fo~ of ~c~c. ~is abiH~ of ~ h~m ~d ~ ~tcm to adapt to ch~ con~fi~ ~ ~ when one ~~s ~e ~ ~ ~ ~ no~mo~ W ~o~. People who smo~ 4 T102322207

the o~yg~-cm'ryi~ c~pacity of blood ~e of the carbon ~ in the smog) m~d by incre~_sing the demands on the heart muscle itseN becm~se ~/co "tine ~ heart r~e m~d makes blood vessels get ~er (v~soconstricffon), thereby incrensing blood pressure. These ~ changes increase the amount of v~rk that the heart has to do per mifiute while, at the same time, reducing the available supply of o~ygen to the heart. Nicotine also interferes with th~ normal reflex control of the heart (12). Compounds in the smoke alter the chemical and physical characteristics of blood. The cardiovascular system therefore undergoes changes to attempt to compensate for all the deleterious effects of smoking. Nonsmokers, however, do not have the ~benefit" of ~ adaptation, so the effects of passive smoking on nonsmokers are much greater than on smokers. This is probably for two reasons: first, nonsmokers' hearts and vascular systems have not attempted to adapt to the acute poisoning from the chemicals in the secondhand smoke. Second, it appears that the cardiovascular system is ex~emely sensitive to many of the chemicals in secondhand smoke and that smokers may have achieved the maximum response poss~le to at least some of the toxins in the smoke, so the small additional exposures a~ociated with passive smoking have ~ktle or no effect on habitual smokers because the additional dose of these toxins is ~mall c~>mpared with what the smoker normally receives. These two fac~ make it imperative to consider the effects of environmental tobacco smoke on the cardio~asc~lar system of passive smokers separately from the effec~_s on active smokers. The quz]~tative differences between the effects of ETS on smokers and nonsmokers explains the relatively high relative ~ a~ociated with passive smoking. compared to active smoking, even though passive smokers absorb much ~er doses of ~e T102322208

tobacco industry i~ f~nd of saying that even a heavily exposed passive smoker only breathes in the equivalent of about of 1 cigarette per day. Leaving aside the philosophical question of whether anyone ought to be required to breathe one cigarette a day under any : circumstances, the smoke from one cigarette is enough to produce substantial effects on the cardiovascular system. The Use of Animal Models in Ca~ovascular Research Animal studies complement human data bemuse it is often not poss~le to do the necessary studies to define biological mech~i~ms of disease in humans f~r ethical reasons. To get around this problem, scientists develop animal models of disease that mimic human disease. Over the years~ these animal models become well characterized and add significantly to our knowledge of the mechanisms and treatment of a wide ~riety of inuess. For example, rabbits and rats are widely used models f~r studying cellular p~ in the heart, as well as the development and modification of athemsclerosis (13). Birds are used to study athemsclerosis. Dogs are used to study mechanical function of the heart. The availability of these animal models has conm'buted significantly to our understanding of the precise mechanisms by which passive smoking damages the heart. Carbon Monc0dde and Ox~_ gen Delivery The fact that passive smoking produces immediate effect.s in nonsmokers, howeve~ significantly strengthens the scientific case that passive smoking causes heart disease. Passive smoking reduces the ability of the blood to deliver mTgen to ~e heart ~e. TI023222.09

o° ~ the ~ygem, thus mdccing the m3,gen carrying capacity of the blood (14-18). If the ca~bo~yhemoglobin (amotmt of carbon moncmide bouml to the blood) gets high cntragh, it can effectively suffocate the individual and prove fatal. Carbon mona~ide poisoning was the o "nginal method of suicide attempted by the lawyer in the opetting sequence of The Cli~nt. Even at the lower levels of carboxyhemoglobin observed in passive smoke~.'it can have an effect on exercise performance and how the heartbeats, in people who already have heart disease. Children of smoking parents have elevated levels of a chemical called 2,3- diphosphoglycerate (DPG). This chemical appears in red blood cells in an effort m modify how they handle oxygen to compensate for chronic ox'ygen staneation (9,16,19-21). When considering the impact of carbon monoxide in ETS on the heart, it is important to remember that ETS is not the only source of carbon monoxide. Given that carbon monoxide has a long half-life in blood, the effects of ETS will add to the carbon monoxide from other sources, and can provide the marginal increase necessary to precipitate adverse effects on the cardiovascular system. Oxygen. Processing in Cells There is also direct evidence from animal studies showing that passive smoking reduces the ability of the heart muscle to convert the oxygen which gets to the heart into the Nenergy molecule~ adenosine triphosphate (ATP), which is like a little chemical battery that is then used by the cell to power the energy-requiring processes. The body needs oxygen because elements within cells called mitochondria take this c~tygen and through a chain of chemical reactiom convert it into ATP. In heart muscle, the ATP is particularly important because it fuels contraction of the muscle (which is n~ for the heart to be.m) as well as the che~ pumps which move calcium and other ions atmmd the mmcie 7 T102322210

mi~ are h3~ an assembly line in which o~'ygen l~oe-s in cm~ end ~ ~ ~ ~t Sm~ h~ ~ ~t ~ese ~ ~ n~ ~rk ~ ~~y ~ ~i~ ~sed ~n ~mp~d m ~e ~e~ of a rabbit. E~n a s~e ~s~ ~ en~en~ tobac~ ~o~ ~du~d ~e ra~ at w~ch ~en w~ ~n~ned ~m ~. F~er ~e~en~ us~ rabbi~ h~ ~ed ~amrs to ~late w~ step ~ ~e ch~ of ~om ~ ~ed ~ ~e E~. ~e e~e ~c~me ~d~e ~bited a ~% ~du~on ~ i~ a~ (i.e. e~en~) ~er a s~e 3@~ute ~s~ to secondh~d ~o~, ~d ~e a~w ~n~ued to ~p ~ pmlo~ed ~s~e. ~er ei~t ~e~ of 3~ute a ~ e~ i~ ~ ~ ~t ~ ~ (~). ~ ~t o~ d~s ~a~ seconded ~o~ ~du~ ~e ~M~ of ~e b~ m get ~n ~gen it does get (~). E~ on ~e~se ~ese bioche~ e~ o~ ~n~d smo~ e~u~ (toge~er ~ o~er ch~ges d~ed later) help e~l~ ~e ~ ~at people s~ly ~ot ~ ~ ~. when ~osed to se~n~d ~o~ (18~). ~e ~c e~ depend on me~u~men~ berg ~ed ~d whe~er ~e subje~ ~ no~ he~ ~ople or people e~e ~ long or ~ach ~ ~ a le~l of ~ ~er b~ ~~d smo~ T102322211

pe~ level of exertion during eaercise, the rn~a,~m]lrn heart rate, and carbon dioxide output. Other studies compardng the ~ ability of healthy individuals to people with heart disease f~tmd that exposure to secondhand smoke lengthened the time to recover resting heart rate at the end of exercise, to the point that the normal healthy individuals took as long as people with heart disease to recover their resting heart rate following exercise. Patients exposed to secondhand smoke who have preexisting myocardial infarction have more ventricular arrhythnfias (irregular heart beats) during passive smoke exposure than when breathing clean air. Passive smoking also significantly increased the amount of a chemical called lactate in venous blood, which indicates that during passive smoking the heart was switching to a greater reliance on anaerobic (i.e., not oxygen requiring) metabolism (27). This finding means that the circulatory system was not able to deriver enough oxygen to the heart muscle to fully meet its demands. The combined effects of reduced cgygen carrying capacity and increased lactate led to less mechanical power being developed during the exercise as well as shbrter duration of exercise before the person reached exhaustion. Even the relatively low (by comparison with smokers) levels of carbcayhemoglobin observed in passive smoke~s may be important because the body normally extracts more than 90% of the oxygen from the blood during exercise, so even small reductions in the efficiency of the caygen transport system can a~ct exercising in nonsmokers (28). These negative eflY.~ act ~tically with the reduced efficiency with which the hea~ muscle conve~.s the vaygen it does received into the ~ molecule ATP beca~__~ of the zztiom of e~avcmm~al tobacco ~taoke o~ 9 T102322212

P1me]ets Secondhand cigarette smok~ affects blood platelcts, in a way which incma~s the likelihood of formation of a thrombus (blood clot). Platclets play an important role in blood dotti~ when they are activated and become sticky in re~spons¢ to a cut ~en blood platelets aggregate inappropriately and form a thrombus, they can precipitate a ~al infarction. In addition, when platelets are activated, the platelet activation factor itself can damage the ~g of the coronary arteries and facilitate the development of atherosderotic lesions (29-34). Platelet deposition in the coronaries and formation of a thrombus can contribute to the growth and progression of an atherosclerotic plaque. Both active and passive smoking increase the aggregation of platelets and thus the likelihood of thrombus formation at myocardial infarction and, through platelet activation, they increase the chances of developing atherosclerosis. Several studies in which nonsmokers are exposed to re~li~ti¢ levels of environmental tobacco smoke, such as levels experienced while sitting in a hospital waiting room for fifteen to t~venty minutes while people arc smoking, show that ETS can increase platelet aggregability to levels approaching those observed in smokers. Large platelets and mean platelet volume are independent risk factors for recurrent or more serious m~nm~lial infarction (35). These data support the epidemiological evidence that passive smoking is a risk factor for both fatal and nonfatal heart disease. In one experiment, nonsmokers and smokers were asked to smoke two cigarettes (32). The smokers" platelets, which were "stickier" than the nonsmokers platelets at the beginning of the experiment, did not s~canfly ~ activity in response to the two cigarettes.. Most ~ the smokers' ~|e~s were ~ ~ becm~ of the 10 T!02322213

(~ with what a smolrer receives on an ~ basis) amoum of trains in cigarettes had no additi~mal effe~. In contrast, in nonsmokers smoking j~st two ciga~ttes significantly increased platelet activity, to the point that it was not signifi_~ntly different from a habitual smoker. This situation illnstrates the fact that the responses of nonsmokers and smokers to the toxins in the cigarette smoke aze often quite different. Based on the data from smoke~s, one would conclude that additional exposure had no effect, whereas looking at the effect on nonsmokers leads to the conclusion that it has a major effect. More important, the same investigator (32) measured platelet activity in smokers and nonsmokers before and after they sat in a room for 20 minutes where cigarettes had been smoked just before the experimental subjects sat in the room. Again, there was no significant change in the platelet activity among the smokers, but a significant increase in platelet sticldness was seen among the nonsmokers to the point that their platelet activation was not discernably different from the smokers. These data, together with the results of other human experiments (30,36-39), indicate that nonsmokers are much more sensitive to secondhand smoke than smokers and that very low levels of ETS exposure can have major impact on nonsmokers' platelet activity. It also appears that the process saturates at low dbses: once'the.nonsmoker has been exposed to even a low dose of secondhand smoke, the platelets are maximally activated much like that to of a habitual smoker so that additional exposure does not increase the effect. These data also indicate that dose-based extrapolations from smokers to nonsmokers using cigarette equivalents" will grossly underestimate the risks to nonsmokers of breathing secondhand smoke. data also support this conclusion. In our stu~es of the effects of passive smoldng ¢m he.art disease, we have ~ that b~ time (another measme of platelet a : vity) is sign camb' shortened more phteL ) in both 11 T102322214

and.rats (42) eaposed at even the lowest doses of secondhand smote, with no additioml effects at higher doses. At a b'mch~ level, studies of cigarette smoke extract on the effects of platelet activity suggest that the tcains in the ¢i~ smoke increa~_se platelet activation factor by interfering with the activity of the plasma enzyme platelet activating factor acetylhydrolase -" (PAF-AH) (43). PAF-AH reduces platelet activity by neutralizing platelet activating factor. Because toxins in the cigarette smoke appear to reduce the effectiveness of PAF-AH in neutralizing platelet activating factor, these toxins may contribute to an increase in platelet activity. The nicotine in the smoke does not appear to be the active agent, but rather some other as yet undefined element in the cigarette smoke (37,43). This biochemical result is reinforced by clinical studies which find that smokers treated with nicotine patches show f~wer changes in platelet activity than continuing smokers despite having similar nicotine levels (44). Atherosderosis In addition to short term toxicity of cigarette smoke, there are long-term permanent effects. In particular, smoking contributes the development of narrowing and blockage of the coronary arteries (atherosclemsis). When someone has a heart attack, one or more of the arteries delivering blood to the heart has become completely blocked and the flow of blood - and hence oxygen - to the hear~ muscle stops. If this flow is not restored promptly, the muscle dies, causing a so-called myocardial infarction. The situation can lead to irregular heart beats (arrhythmias) which can also be fataL The coronary arteries are blocked through a proce~ ~ in ~ ways to blockage of a pipe in ~ p.~ The ~ of the ~ is fast ~ ¢Rher chemically T102322215

hardezfing aud ~ lmown as a~. As the vessel mrro~ the ~s ~e that it will become completely blocked and cau~ a heart attack, eithe~ through slow due to plaque or, more ohen, a blood dot lodging in the small remaining opening. The toxic chemicals in the cigare~ tte smoke, and particularly polyvyelic aromatic hydrocarbons (PAHs), f~:ilitatv ~ pro~s~ both by d ~ama~n~_g the cells ~ the coronary arteries and also by stimulating plaque growth through a process similar to that of a benign tumor. In addition to their role in acute thrombus f~rmafion, platelets are also important in the development of atherosderosis. Once then is damage to the arterial endothelium (that is, the lining of the coronary arteries), either through mechanical or chemical factors, platelets interact with or adhere to the arterial wall and initiate a sequence that leads to formation of atherosclerotic plaque. Ifplatelet aggregation is increased because of exposure to ETS, the chances of platelets building up at that endothelial injury site will b¢ increased. Thus, in addition to contributing to short-term effects by increasing the likelihood of thrombus formation, the effects of ETS on platelets also increase the chances that endothelial injury will lead to arterial plaque formation and atherosclerosis. ETS also plays a role in causing damage to the endothelium and initiating the atherosclerotic process. Experiments in humans have indicated that ¢v~n short-term exposur~ to ETS - like active smoking (45) - significantly increases the appearance of anuclear endothelial cell carcasses in the blood of people exposed to ETS (or other tobacco products) constituents (30). The appearance of these cell carcasses indicates damage to the endothelium, which is the initiating step in the atherosclerofic p~ The appearance of endothelial cells after passivv smoking i~ nonsmok~ aher just 20 ~tcs a hospital waiting room, is almost as g~at as in ~ smoking in nWL~~ L~ T102322216

~es to the ~e ~j~ ~a ~d promote ~i~ smo~ bo~ ~o~ adol¢~n~ whose p~n~ ~o~ ~ ~o ~ .~ ~r~ ~ pla~s wb¢m ~o~ ~ ~d ¢~bit l~r I¢~ of ~ (~¢d good cholesterol) demi~ ~popmte~ ~ people bma~ d¢~ ~ ~¢r adj~ ~r g~der ~d s~. ~ ~ ~o ~m~s ~¢ ~ of de~Iop~ ~m~ he~ ~e. ~e Role of ~v~]ic ~mafic Hy~bo~ ~) M~ a~emsdemdc pla~es (46). ~r~ 7,12-~Ib~m(~h)~ea~ ~) ~d ~(a)~a~ ~), developmem of a~emsdems~ (47-51). ~ ~ m ~~t ~fiment of E~ (15~2). Chmges • ~ inj~ ~at loa~ to p~t¢let a~gafion ~d pl~u¢ Se~r~ s~s ~ ~di~ted ~at pol~c ~fic ~~m pm~mn~y. co~e~ additio~ of ~olesteml, which m~ ~tate ~co~orafion of d~g ~ a~d~c pm~sa. T102322217

- Se~,ml eart~ ~ smdi~s de~monstrated that it was possx'bte to acce~ate ~ athe~sd~tic proce~ in experimental ~ (warious spedes of ~ and pig~om) by injection of BAP and other related PAHs (49.50). In additioa, it was possible to extract cancer-like cells from plaques, which cau then be transplauted to another ~imal and produce similar cells (50). ." In addition to ~ biochemical evidence examining the effects of specific components in ETS on the development of atherosclerotic lesions at a cellular and molecular level, there have been a series of animal experiments which demonstrated that short-term exposure to environmental tobacco smoke dramatically increases the rate in which lipids are deposited in arterial linings. This is the initial process in the development of athexosclerotic lesions. We (40) exposed three groups of rabbits on a high cholesterol diet to ten weeks of exposure to secondhand smoke from Marlboro cigarettes. (This rabbit model has been used to study atherosclerosis since 1908 (13).) The animals were exposed to the secondhand smoke six hours a day, five days a week for just ten weeks. One group was ¢xposod to smoke at levels that would be observed in a smoky bar or the others were exposed to levels about three times as high. The high dose group was exposed to pollution levels comparable to thbse observed in a Mazda 626 with the windows rolled up at 4 cigarettes/hr being smoked (53). With just ten weeks of exposure (a total of 300 hours), the fraction of pulmonary artery and aorta covered with lipid deposits nearly doubled compared to control rabbits who ate the same diet but were not exposed to ETS. This is a short term exl~sure, even fvr a rabbit. This effect appears to be directly due to elements in the cig ~arette smoke it~.lf, rather than a nervous ~ to berg ezposed to the smoke ~ ~ have ~ the T102322218

active smokers.) To ad&ress ~ question, w¢ (41) exposed rabbits in an ¢ape~cnt ~ar to that just described to second-hand smoke but gave half the rabbits the tmta-blocking drug Metropolol, which block~..s the effects of catecholamines. As expected, the anlm~i~ receiving the drug developed fewer lipid deposits than those who w~re receiving a placebo (saline), but this effect was independent of whether the rabbits were breathing secondhand smoke. Therefore, the secondhand smoke effects on the development of atherosclerotic-like lesions in the arteries was not mediated by the increased levels of catecholamines. One criticism that might be raised of this study is that the rabbits were on a high cholesterol diet (54). This experimental model of atherosclemsis, which has been used since early in the century, requires the rabbits to have a high cholesterol diet in order to develop any lesions within a reasonable length of time. Other investigators (55,56) used cockerels (another standard animal model for studying atherosclemsis) exposed to secondhand smoke showed similar increases in the amount of plaque developed in young cockerels (between 6 and 22 weeks old) who were exposed to secondhand smoke six hours a day, five days a week for twelve weeks while on a low cholesterol diet. These chickens were exposed to lower levels of secondhand smoke than the rabbits studied and were eating a normal, low cholesterol, diet. While there was no difgereuce in the plaque incidence between the cockerels breathing secondhand smoke and the cockerels breathiug clean air, there was a significant acceleration in the ~ of these plaques, in a dose-dependent ~er in the ETS-exposed bi~s. The ~gens in the smoke appear to be acting as a promoter to facilitate the development of plaques, raflaer than i~itiate that these,effects are due to the carbon mon~de in the smoke, 16 T102322219

Even so, exposure to secondhand smoke f~ a relatively brief time (con~sponding to about 0.4% of their life span) si/gtificantly accelezated the development of pla~ues. It is reasonable to conclude f~om these results that in tkis animal model, moderate exposure to secondhand smok'~ for a brief pe~od in early life is sufficient to markedly accelerate the development of atherosclemtic plaques. The exposure levels used in ~ experknents axe comparable to those observed in many workplace and household environments. The fact'that it is possible to/nduce atherosclerotic-like changes /n two different species of experimental animals with only a few weeks' exposure to secondhand smoke similar to that experienced by people in normal day to day life, is a very important finding linking the epidemiological and biochemical evidence that passive smoking causes heart disease. On one hand, the epidemiological studies show an increased risk of heart disease on a population basis among people exposed to secondhand smoke. On the other hand, there are biochemical studies showing that elements in the smoke, particularly BAP and the other PAHs, can produce cellular changes that occur in atherosclemsis. The experimental s/udies on rabbits and cockerels, which do not suffer from the potential confounding variables in epidemiological studies, bridge this gap by showing that it is possible to induce atherosdemsis in experimental animals with ETS. Finally, there are also dam in humans showing that passive smokers have f~nificanfly thicker carotid artery walls (the artery in the neck) than people who never were exposed to passive or active smoking, with a dose- response relationship (58). the animal ~ents. These results are consistent with what ~ ,,~yald expect The causal link between passive smoki~ and atherosclexosis 17 T102322220

Fr~ radicals in ETS and ig~hcmic dmna~_ As already discussed, ~ ~ s~ ~de~l~ m s~~d ~ ~¢fien~ ~ a ~mr ~o~ ~ dea~ ~ ~ h~ ~ ~d ~o non-fa~ o~n~ ~ ~ ~~ new e~den~ ~at p~ smo~ ~e~ ~e out~ of ~ ~c ~nt ~ ~e he~ ~u~ ~e ~W of ~ ra~. F~e r~ ~ ~y ~a~ ~gen p~u~ w~ch ~ e~mely deserve m ~e he~ m~de ~11 mmbr~e ce~ (59,60). ~e r~ h~ been ~pfi~ted ~ ~r ~ ~ ~ he~ ~e~e.) By dis~p~g ~e a~W of ~e ~mbr~es ~ ~e he~ m~de ~ ~e ~ ~y ~teffem ~ ~n~o~ of ~e he~ m~e. ~cy h~ ~en ~y ~fi~ted d~g ~e he~ h ~ what h ~ ~ ~pe~on ~j~." ~~ion ~j~ when bl~ st~ fl~g m he~ m~e ~er it h~ ~en ~m~pt~. ~ ~e bl~ ~ st~ped, ~e he~ m~le ~ de~ of ~ge~ a si~fion ~ ~e~ For ~ple, a re~sion inj~ ~ oc~ when ~meone s~en a he~ a~ ~e~), fog~d by ~a~ent ~ ~opl~ (fo~ ~e ~e~ o~n ~ a b~n) or a clot-bus~g ~ag ~at ~solms ~e clot (t~mb~) bl~ ~e ~m~ ~e~. Repe~ion ~j~ ~ ~o o~ when a patient h~ had o~n h~ ~ ~ ~ of blood m ~e he~ ~ s~ ~en s~ ~ ~ea he~ muscle ~ depfi~d of ~gen ~u~ the ~lls which s~e ~e radi~ ~ depleted. ~e ~ ~ n~y ~d ~ ~e ~R ~ it r~~ ~d ~d~ ~ no~ T102322221

Studies of humans aud several species of ~ indicate that low ezposums to nicotine or othar cigarette .~moke constituents significantly worsen repeffusion injury. For example, slowly infusing the nicotine of just one cigarette doubled the effect of the reperfusion injury on heart muscle of dogs (61). "Ibis is so'low a dose of nicotine that it had no effect on heart rote, blood pressure, shortening contraction of the heart muscle itself as the heart beats, or other hemodynamic measures of cardiac function commonly affected by nicotine in active and passive smokers. After an ischemic episode in which the blood supply to a section of the heart was interrupted for fifteen minutes, the shortening of the nmscle in the heart wall during repeffusion was reduced to 50% of the pre-ischemic values. When the dog was exposed to the nicotine from just a single cigarette, the muscle lengthened to only 25% of control values. Thus, exposure to a very low dose of nicotine doubled the impact of the reperfusion injury. When the dog was given a free radical scavenger which neutralized the free radicals due to the nicotine, this effect was obliterated. The effects of free radicals induced by passive smoking have been explored at the cellular level (62,63). Rats who were exposed to secondhand smoke from two cigarettes a day for two months exhibited severely damaged mitochondrial function during reperfusion injury, so that the ability of cardiac mitochondrial cells to convert c0cygen into ATP was much more compromised during ~perfusion injury among rats exposed to these low doses of secondhand smoke than among control rats This is another way in which the tc~dns in the secondhand smoke can interfere with energy metabolism in the cell. TI023~

control free radicals (64). In addition, passive smoking by h~ sensitizes lung neutrophlt~ (65). As with plate.lets, neutrop~ are an important element of the body's defenses ~ ~ction and damage. Inappropriately activated neutrophils, howevei; release oxidants and these elements can play a role in tissue damage in p~ssive smokers. In a group of passive smokers exposed to just three hours of sidestream smoke, there w~re significant increases in the circulating leukocyte counts and stimulated neutrophil migration. Like the other respomes, the responses to the exposure to secondhand smoky were greater in nonsmokers than in smokers, again suggesting that the biochemistry of ETS in passive smokers is different than in active smokers, with the passive smokers being more sensitive to these elements. Likewise, when hamsters were exposed to the ETS from six cigarette a day for eight weeks, the activity of anti-cmidant enzymes in their lungs nearly doubled. These enzymes are natural factors which act to reduce the activity of free radicals. This study deals with neutrophils in the lungs, but it is reasonable to assume that the neutrophils exhibit similar effects throughout the body, since they are transported by the blood. Myo~rdial Infarction (Hev-rt Attack) There are also direct animal data to show that secondhand smoke promotes more tissue damage following myocardial infarction (heart attack). Dogs exposed to secondhand smoke one hour daily for t~ days, then subjected to blockage of a coronary artery developed myocardial infarctions (dead tissue) that were twice as lmge as those of control dogs who breathed clean air (66). T102322223

.. We(42) condnoedancxperimentinrats~~totherab~~~ above, to investig~e the effects of ETS exposure on infarct size. Rats were exposed to secondhand smoke six hours a day for three days, three weeks or six weeks at concentrations similar to a smoky bar or Mazda 626 car with a smoker inside (53). We tied off one coronary artery for 35 minutes and then released the tie to reperfuse the tissue. W~'found a dose-dependent increase in infarct size with the longest exposure (180 hours total exposure) yielding infarcts that were nearly twice as laxge as those of the control group that breathed clean air. The facts that these effects could be induced in experimental animals and that we observed a dose response relationship axe strong evidence that the passive smoking caused the increase in infarct size. In addition, there was some elevation of infarct size at even the shortest durations of exposure, suggesting that any exposure would worsen a myocardial infarction. There is no evidence of a threshold effect. Although smokers seem to be less sensitive to effects of passive smoking than nonsmokers, it is important to recognize that even k~v doses of cigarette smoke can have important effects for smokers. For patients with coronary artery disease, smoking one cigarette s!gnificantly increases the coronary vascular resistance (67). Thus, at a time when demahds for oxygen and blood supply to the heart are increasing (12,68), even a single cigarette can dramatically reduce the ability of the coronaries to transmit blood. In addition, in habitual smokers smoking a single cigarette causes an increase in the stiffness of coronary arterial walls, and this increased stiffness may be related to the rupturing of the atherosclerotic plaque which can be an important element in myoca~al infarction (69). It is likely that ~ smoke ~ produce fnnilar effects in nonsmokers. 21 T102322224

an~ has n~wr smok~ n~r be~n e~sed to ~ smo~ and ~y d~ed a~~ habit. ~e~ ~m s~est ~at ff ~ ~du~ s~ a he~ a~ w~e ~ or showy ~r ~g ~ to s~n~ ~o~, ~e h~ a~ ~ato~ ~mequen~s of ~ ~ ~ ~ ~~t. ~ ~tc ~at e~on- . te~ ~~s ~ sec~d smo~ ~ ~ d~~. Epide~ol~ Sm~es ~ ~ady not¢~ is~c hem ~¢~¢ disuse mo~W ~ p~si~ smo~ (7 of no~o~ ~men ~¢d to ~n who smo~ (7~78) ~d 5 of no~o~ m~n (7~73-77,79)). ~ o~r sm~ pub~h~ ~ abs~ also sh~d ~¢d ~k (~,81), but ~¢ ~ ~uded ~~ ~ ~ n~r published in ~R.) dis¢~e for no~mo~ ~d to smo~n ~mp~d to no~o~n ~d to nonsmokers, ¢~n di~t, w~igh~ ~d ~¢). h ad~fio~ ¢i~t of ~¢~ sm~¢s ~ a ~ifi~ do~-m~¢ ~lmio~hip in whi~ ~¢r ~s~ m ¢~m~n~ tob~ smo~ ~ ~t~ ~ a higher risk. ~s¢ the risk of he~ ~¢~e d~a~. ~e gen~r~ d~si~ nonsmo~ m~ied to smo~ ~t~ ~ ~ ~~ ~ ~~d~ L0. h11~~~~

nonsmoker married to a smok~ than ff y~m a~ a nonsmoker married to a ncmsmoker. The f~ct that the obseTved risks are of comparable magnitude acr~s studies done in many countries and comro~ for a variety ~ the other risk factors for heart disease sirengthe~s the confidence one can have in reaching a conclusion that passive smoking causes heart : disease. Some of these studies used marriage to a current smoker as the measure of exposure to ETS, whereas others used marriage to an ever-smoker (even if the spouse was currently a nonsmoker) a~ the measure of exposure to ETS. Given the fact that the effects of ETS on the heart decline quickly (particularly when compare to lung cancer) when exposure ends, the design of these studies is often biased against detecting an effect of ETS on heart disease. The problem in interpreting any epidemiological study is that the results depend not only on the t~ue effect of the toxic agent being studied (in this case environmental tobacco smoke) but also on the specific individuais who happen to be selected in the random sample obtained for the study. This uncertainty is commonly quantified using a "confidence ifiterval" which provides a range within which the true relative risk is l~ly to He. The 93% confidence interval covers the range in which one can be 95% confident the true relative risk lies. If the 95% confidence interval excludes 1.0, one can conclude with 95% confidence that passive smoking changes the risk of dying of heart disease. While the question ~f statistical signiticance deals with whether or not the confidence intervaI includes one, the fact is that we can be 95% confideat that the true relative risk ties somewhere in the ~ce ~ It is equa~y ~ mat the true risk could be at the T102322226

co~dence int~vals, w~ can conclude that three of the studies, taken alone, provides enough evidence to b~ 95% certain that passive smoking increases the risk of dying of heart disease. The probability of obtaining three such positive tests by chance is only 0.000125, or:about I in I0,000. These three positive studies alone provide adequate evidence to conclude that there is a link between ETS exposure and death from heau-t cliseas¢. The confidence intervals w~ have been discussing correspond to so-called two-tailed test statistical, in which one considers the possibility that passive smoking increases the risk of heart disease but also considers it equally likely that passive smoking would reduce the risk of heart disease, i.e., breathing second-hand smoke would b~ protective of coronary artery disease. Since no one has asserted that passive smoking is good for people~ headth, a two-tailed test is, in fact, overly conservative. It is more appropriate to use a one-tailed test which tests the question of whether or not passive smoking increases the risk of heart disease. In this case, we would see whether or not the lower bound of the two-tailed 90% confidence interval includes 1.0, because the lower bound of the two-tailed 90% confidence interval is exactly equal to the lower bound for the one-tailed 95% confidence interval. When the lower bound of a two-tailed 95% confidence interval exceeds L0, we are actually 97.5% confident that passive smoking increases the risk of heart disease. Because cigarette smoke is a known toxin, OSHA should not consider the possibility that ETS prote~.s people from heart disease. In terms of the overall risk assessment, OSHA should base its analysis on one-tail tests. Ti02322227

statistical significanc~ would b~ a true negafiv-c, m~-~g ~t ~¢ sm~ d~ ~ ~ ~m~c none ~d. ~c~ ~ a s¢~nd ~n ~at ~ s~~ ~~: ~ ~ple ~. ~ s~ ~ ~t of ~ ~d~rl~ biolo~ ~ab~ ~ ~e p~afion be~ sm~. ~e ~fi~ of ~ s~~ test - ~e c~led p~r of ~e ~t - ~~ ~ ~e s~ of ~e sm~. ~e ~ogo~ si~fion ~m ~fi~l pubic op~on po~ ~ ~t l~er pubfic ~on ~ ~ mo~ p~e ~sul~.) U~o~a~ly, epide~olo~ sm~es ~ ohen ve~ e~e~i~ to ~ndu~ ~ it is d~lt to a~ enou~ ~es to ob~ ~ adequate s~fi~ p~r to be able co~dently ~nclude ~at a f~ to ~h s~fi~ si~~ me~ ~at ~ ~ not a ~e Just ~ it ~ ~do~ ~de~d de~le m ~ able si~fi~ce on a ~sifi~ ~ncl~ion ~ 95~ co~den~, it b ~de~d de.able to ~ 80% p~er ~ a s~fi~ test to co~dendy ~ach a nega~ ~nd~ion ~m smt~ti~ ~. U~~tely, be~e of s~ s~ple ~ n~e of p~ive smo~ ~d he~ ~e~e ~ach ~ ie~l of p~r. ~dee~ m~t of ~c ~es h~ve p~ers bel~ 10~ (m dete~ a 20~ ~e~e ~ ~k ~d~ted ~ p~ s~) (3). ~at ~ me~ b that i~ ~ ~ p~sive smo~ ~ed ~e ~k of h~ death by 20~, most of ~e epide~olo~c s~es ~d h~ le~ ~ a 10% ch~ of ~u~ly ~po~ a smt~d~y si~fi~t effe~ ~s 1~ p~r ~ it ~~le to ~ach a ne~d~ ~nd~ion ~ed ~ ~ of ~ese sm~, ~ ~e ~t, ~e~ ~.d ~ do, p~~ ~n ~t 11 ~ ~ ~ T102322228

~ ~c~e t~t t~e ~tu~l studies)? Ttze ~s ~¢ ~ ~ ~e ~. P~ ~ of ~e ~es ~ a s~ed mem-~T~ not ~y h~ ~e ~~e of ~~ ~e ~ of ~e • = one co~d much mbm pm~ely es~m ~e a~ ~ ff ~, ~at~ ~ p~ smo~g. P~g ~e res~ ~ ~ ~ sm~es ~el~ ~ es~te of ~e ~lafi~ ~ for ~ of he~ ~e~e of ~ (~ a 95% ~den~ ~te~ ~en~g ~m 1.1 to 1.4). ~e~, ~ ~ be morn ~ 95~ ~dent ~at p~i~ smo~ ~m he~ dhe~e ~d more ~ 97~ co.dent ~at p~i~ smo~ of dea~ ~m he~ ~e~e. ~en j~t ~e 5 ~st of ~e sm~es for he~ ~ dea~ ~ ~le~ one ob~ ~ e~n higher ad~t~ mlafi~ ~ ~r p~ ~o~ ~d h~ ~e mo~y of (~ a 95% ~dence ~te~ ~en~ ~m 1~ to ~). taming 7 l~r qu~ sm~es (de~ed ~ not ~n~g ~r ~ables) I~ ~e pooled es~te of ~e mlafi~ ~k h a m~ ~tems~ ~ be~e it ~di~tes ~at ~¢ be~er ~e qu~ of ~e study - de~ed ~ ~n~ ~r o~er ~o~ ~n~bute W he~ ~e - ~e hi~er ~e ~ a~buted to s~n~d smog. ~ o~er. ~s, ~e co~o~g fa=ors ~ a~y m~k ~ toba~ ~d~, w~ch ~fi~s studies for ~ to ac~t ~r ~o~ ~ables, o~n ~e~n ~at f~ w a~t ~ ~de~ ~~y ~t~ ~k ~d ~s it ~ ~at E~ ~ ~m d~~ ~ it m~ h. ~e ~ sima~ ~ ~t ~e TI02322229

analysis, particularly within the context of their advertising campaigns directed at discrediting the EPA risk assessment of passive smoking and lung cance~ The tobacco industry suggests that such an analysis is somehow controversial or ~uuscientific.~ ~ is a si~ ~ f~om both scientific and common sensical points of view. The process of p6oIing epidemiological studies ~and stratifying across studies is well-established in epidemiology. More important, it is simply good judgement to take into account all available information in making any decision. To understand ~ point, .consider the following simple example: Suppose I have a coin and offer you the opportunity to make abeL If the coin comes up heads, I will pay you $10; ff it comes up tails, you pay me $5. This seems like a good bet for you, so you agree. I flip the coin and it comes up tails; you give me $5. I then ask you would you like to try the bet again; you say yes and I flip the coin again. It comes up tails and you give me another $5. I ask ff you want to play again, you say sure, and I flip the coin, it comes up tails, and you give me yet another $5. At this point you may begin to become suspicious and say it does not seem very likely that I'would get tails three times in a row on a fair coin, which should have a 50/50 chance of coming up heads or tails on each flip. You just did a recta-analysis. I respond to you, you can not do a recta-analysis because it is conlroversial. I will only permit you to draw conclusions on the data fzom one flip at a time, without using information from the other flips, even though one flip of a coin does not contain eno-agh information to determine whether the coin is faix. In statistical paxlance, a single flip of a coin is.not a ~ eam.~..gh sample size to have adequate ~ to be 95% ccmfide~t that T102322230

you.wishtoma~anott~b~t? Stnc~youcaaoalycondd~t~la~-tf~p, yoa~,y~s, lflip ~ ~ ~ ~ ~p ~ ~ ~ ~s ~d it ~m~ up ~ 11 ~. ~ ~ ~d~ ~ ~p ~ ~fio~ ~u near ~t ~o~ ~fion to ~n~ud~ ~t ~¢ ~ ~ not f~ e~ ~ ~ ~o~ble pe~n ~d.s~ ~at a 11 ~es ~ not f~. Does ~ m~ it h ~ible w ~p a ~ I2 ~ ~d get ~ 11 t~es? ~solutely no~ but it ~ ~ ~ly. ~e pmbab~ of ~pp~ a ~ ~ 12 ~es ~d get~ 11 hea~ ~ o~y .~ or about 3 ~d 1~, not a ~ ~ly out,me. Gi~n ~e 1~ ~bab~ of ~ oc~ ~ ~ ~ ~ ~e ~ w~ f~, most ~ople ~uld ~je~ ~e ~e~on of a f~ ~ ~d ~n~ude ~t ~e ~ h not ~. No r~onable pe~n ~d ~je~ ~ mem-~h. ~o~er obje~on conduced on p~i~ pooled. ~ ~d flipping g~e ~m~e some~es I ~d a ~e ~d ~me~es I ~p~d a qu~ when I prided bo~ ¢o~) It ~ ~e ~at no ~ sm~es ~ idenO~ - ~me ~ pms~ and others ~ed ~e-con~l me~olo~ ~¢y diffe~nt sm~es ¢on~Bed ~r ~nt ~~g ~bl~ ~ ~nt ~ ~e de~t with men ~d some d~t broad outl~es ~d go~, h~r, ~ ~ ~es ~ ~. ~de~ ~n ~¢ ~la~ly minor diffe~n~s helen sm~es, dise~e demh. Mo~, T102322231

~ ~ all ~ from the same limiu~iou~ au~ po~'b~e co~ ~~. ~e ~¢ ~fion ~ ~y ~g~ when ~de~ ~¢ sm~ of p~ ~o~ ~d d~ ~m ~c he~ ~e~e. ~e~ ~ 12 ~¢s (~ ~n ~ ~mb~ed) ~d 11 of ~em ~ ~ elation ~ ~ ~ p~ smo~ h~ no ~ on ~e ~~ of ~ ~m ~che~c he~ ~, a~ut h~ of ~ese sm~es ~o~d h~ sh~ ~ elation ~ ~k (mlafi~ ~ a~ LO) ~ a~ut h~ of ~em ~o~d ~ sh~ a mdu~on of ~k ~o~ated ~ p~ ~o~g (mla~ ~ ~l~ 1.0). ~em is o~y a 0.~ ~mb~ of ge~ 11 out of ~ sm~es sh~ ~ ~e~fion of ~ ~ ch~. ~ j~t ~ ~ ~e ~ ~ple ~~ a~, ~ ~je~ ~e ~se~on ~t p~sive smo~ ~ not ~lated m ~e dea~ ~m he~ ~e ~d conclude ~at it ~. ~e o~y ~ren~ helen ~ ~ ~d on a ~ple ~ ~ me~phor, ~d ~e me~-~ ~i~y done ~ epide~olo~ ~ ~t ~e epide~olo~ pmced~ ~ ~fiy ~r ~n~s ~ ~ple s~ ~d ~e ~ o~d ~ ~m • e indi~du~ s~e~ w~ch ~m ~ mo~ p~e es~ of ~e ~led ~lafi~ ~k ~d ~so~ated 95% co~den~ ~te~. ~e pmced~ ~at ~e toba~o ~d~ a~ of c~ide~ e~h study ~ ~olafion ~s no mo~ sere ~at ~e ~ g~e ~ wH~ ~ i~o~d ~e fa~ ~at ~u lost e~ ~p. ~o 11 s~es (~~ endp~ ~r men ~d ~men sep~a~ly) ~at ~ed n~-~ ~ac d~e~e ~dpo~, such ~ a non-~ ~~ ~o~ p~sen~ of ~ or el~n ~k~ E~ ~~ ~e~~~ob~9s~~ ~ T102322232

confidence interval of 1.1 to 1.6). Three of thes~ studies show a dc~se-response relationship, with higher exposv_r~s of s¢coIldhand smoke be/ng associated with larger increases in risk. The fact that passive smok/ng increases the risk of non-fatal corcma~ events as well a~ iatal coronary events is consistemt with what we know about the physiology and biochemistry of the effects of passiv~ smoking on th~ hea~. The only differemce between a fatal and non-fatal coronary event is th~ severity of the event, i.e. whether or not the insult to the heart is so large that the cardi'ovascular system~ natural adaptive defenses can compensate for it. If the cardiovascular system can compensate, or the individual reaches medical care quickly enough, the coronary event will not be fatal. If, however, the insult to the cardiovascular system is too la~e for the body to compensate or, alternatively, ff the individual does not reach appropriate medical care quickly enough, the event could become fataL Thus, the presence of an effect of passive smoking on both fatal and non-fatal coronary events is an additional important piece of evidence supporting the fact that passive smoking causes heart disease. l~blicad~n Bi~ Publication bias ~s usually defined as a propensity not to publish papers that fail to reach so-ca!led statistical significant. A common claim made by the tobacco industry is that there is a publication bias in the literature against negative studies on the health effects of passive smoking, and this pubiica~on bias explaim why there is such a consistent picture implicating passive smoking a~ caus'mg ~o.rious ~d~ases. ~ ~ was made prominently T102322233

.- W¢ braze acldmssed the qc~stion of ~ or not them is ¢vidcace f~r a trablication bias in the area of the health ~ of ¢nv;=~onmental tobacco smoke by careflzlly ¢~mi~i~l,~ the literature on passive smoking and lung cancer (88). We included both the peer reviewed literature and the non-peer reviewed literature which appeared in tobacco industry sponsored symposia on environmental tobacco smoke. We could find no evidence of a publication bias in the area of enviromental tobacco moke and lung cancer. Indeed, partially due to the presence of so many tobacco industry-funded symposia, there might be a bias in favor of publishing negative studies (89-91). While we have not done a similar study of passive smoking and heart disease, it is doubtful that there is any publication bias in this area either. As discussed above, the majority of the epideminlogical studies, taken one at a time, do not reach statistical significance. Thus, if one simply examines the literature in the area of passive smoking and heart disease, the fact that most of the studies are not individually statistically significant argues strongly against the existence of a publication bias against negative (non-significant) studies. Moreover, the only two studies that could be located that had only been published as abstracts (and not gone on to full publication) reported increased risk of heart disease associated with passive smoking. The major source of publication bias is not the peer review process, but dexisions by authors not to submit a paper that fails to reach statistical significance in the first place. Given the level of interest in the issue of environmental tobacco smoke, it is doubtful that people would re~ain f~m publishing a well-done study addressing this question. The only way in which failure to publish certain results could be seriously bias.ing the ~sul*~.s of the o~er~,~ll analysis o~ pas~ smoking and heart~ disease would be ff reseazehers I~e~ere~ly 31 T102322234

(i.e., yields relative risks below LO). No one has ass~rted that this is th~ situation. R~-versibiliw of Effects Like the effects of active smoking, the effects of passive smokin~g on the heart represent a combination of acute toxicity and long-term damage. To the extent that the effects of passive smoking on the heart represent acute toxicity, removal of exposure of the individual to secondhand smoke will result in a commensurate reduction in the risk of adverse health consequences. While there have not been studies of the cardiov~cular effects of removing secondhand smoke exposure f~'om nonsmokers, ~ issue has been studied for active smoking. When an active smoker stops smoking, his or her cardiovascular system is working measurably better within one d~, and the excess risk of heart disease death returns to half that of a nonsmoker in one year (92). Based on this evidence, it is reasonable to conclude that when a nonsmoker ceases being exposed to secondhand cigarette smoke, it is likely that their excess risk of cardiovasctdar events (both f~tal and non-fatal) will resolve unless the secondhand smoke has precipitated an acute cardiac event, which is not revers~le. Selection of Endvoints for the OSHA .I~.'sk.A~sessment Because secondhand smoke can cause fatal as well as non-fatal cardiac events, bo~ of these endpoints should be considered in OSHA~ analysis. This is particularly important because there are approximately three times as many nonfatal cardiac events as there are fatal events. A non-fatal coronary event can have dramatic impacts on the quality of Ii~, health and economic situation of workers, particularly since heart ~ease is very expensive to treat. ~ the strong epidendological and ~ogieal and biochemfical evidence that T102322235

~ smoking c~n have ~ate effec~ on c~diov~cul~ system ~ ~ ~ ~ ~d ~ ~PK M~ ~ pm~s ~ ~e a p~~fi~y b~ ~efic ~BPK) ~ m es~m ~~ ~d e~ of E~. Such a m~el ad~ ~e~s~ ~pl~ ~d ~~~ W ~ ~k ~sessm~nL PBPK m~e~ ~ most ~ when it ~ n¢~ to ~ ~po~fiom ~m ~ ~m of ~# dose e~o~ sm~es of a m to 1~ d~¢ en~men~ or o~pafio~ ~osms or when it ~ nece~ m do ~ss-spe~es e~apola~om ~ pm of a ~k ~¢ssmenL ~¢s¢ simafiom ~¢ when ~e~ ~ not epide~olo#~ dam ~2able ~r h~ ~sed to ~fic 1¢~ of ~¢ t~ ~ quesfiom ~ situation does not e~t ~r E~. ~e ap~ ~ ~¢o~ PBPK m~¢~ ~ ¢v¢~ one of th¢~ p~mete~ ~ a ~tenfi~ ~ of e~n Mo~g ~ ~t of ~¢ ~m • at no~o~ ~nd ~nfly to ~ ~ ~o~n ~nd to ~ ~ ~d p~i~ ~o~g me of &m ~m smo~ to ~ ~e ~ete~ ~ a PBPK model co~d ~duc¢ ~mfi~ e~rs ~ ~ p~om wM~ most ~, ~d ~~fly underrate ~¢ ~k of E~ to nommo~. ~ ~n~t to m~t ~ ~at OS~ mint re#ate, ~¢~ ~ a ~ of dam ~r h~ ~d to ~fic link of E~ ~ ~ ¢n~m~. ~¢ ~k ~m,nt shoed be b~ed on ~e h~ epid¢~olo~. ~ ~labi~ of ~ase dam ~s ~0 ~ m~¢l ~¢~. ~ ~¢~¢nt b~ ~ the hm~ epid~olo~ ~ ~ s~pl¢r ~d m~ ~abl¢. E~n~c I~a~s T102322236

As documented in the May 1994 issue of Co~mer 1~ _.eports ~ "Self-Scrv~ Surveys: The 30 Pe~ent Myth" (93), ~ claim of reduced business, typ~.~lly 30%, is a public relatiom fabrication. It is based on highly selective surveys or careful1~y worded claims of what people "expect' to happen when the ordinances are passed, rather thsn objective data on actual effects. To address the question of what the actual effect of implementing a smoke-free restaurant ordinance is, we (94) obtained the data on total restaurant revenues sales from the California State Board of Equalization and the Colorado Department of Taxation for the 15 cities in the United States which have had 100% smoke-free restaurant ordinances on the books for long enough to obtain at least one year's worth of dam. We also obtained data for several years before the ordinances were passed, so that we could establish any historical trends or seasonal ,,~'iability, as well as get a good estimate of the underlying random fluctuation in restaurant sales. We analyzed these dam in two ways: First, we looked at restaurant sales as a fraction of all retail sales in the 15 communities. Second, we compared sales in communities with ordinances to communities with similar geographical and socioeconomic factors but no ordinances. We concluded that overall there was no effect of these ordinances on restaurant revenues. In particular, in Beverly Hills and Bellflower, Cali~rnia, two c/ties where the tobacco industry successfully repealed the ordinances based on clahns of a 30% d~'~p in business, there was no such drop. Indeed, in Bellower sales were up during the period the ordinance ~ in force. T102322237

resta~arants, ETS levels are 1.6 to 2.0 times as high in restaurants as in office workplaces. Levels in bars were 3.0 to 6.1 times higher than in office workplace~s. ~ incr&tsed ETS exposure is reflected in lung cancer ratesin food-service workers: there is an exces~ lung cancer risk of approximately 50% (range 10% to 90%) among food-service workers compared to the general population, even after controlling for active smoking among these workers. The effects on excess cardiovascular ~_sk would be expected to be similar. For the reasons discussed above, however, the number of cases of fatal and non-fatal heart disease among food-sezvice workers mused by ETS will probably be an order of magnitude greater than the number of lung cancer cases. For this reason, it is imperative that OSHA cover these workers in its rule. Issues of Implementation In response to the evidence that passive smoking is dangerous, a growing number of communities have been enacting legislation restricting smoking in the workplace and in public places. In addition, many businesses have been voluntarily creating smoke-fxee v~orkplaces." While this process has been vigorously opposed by the tobacco industry - often to the point of promoting preemptive state laws and forcing political referenda in attempts to overturn these measures - the rate at which local governments have been acting to control ETS has been accelerating. By August 1994, there were nearly 600 local ordinances in the United States, mandating clean indoor air and restricting the tobacco industry~s access. to youth. Indeed, ~n California, Phih'p Morris tobacco company has alv-~ly spent $2 milIion 35 T102322238

• These local tobacco control ~ whether amndatvd by la~ or sokmmry, have proven to be effective and reasonably self.enforc/ng (96,97). By the time local ~ are enacted, they reflect the popular will. The campaign for enactment, particularly because of the controversy generated by the tobacco industry in opposing the measures, takes a public profile in the community. The resulting "discussion and consemus w~ch emerge play an important role in the effectiveness of these tobacco control activities. In order to be effective, it is important that any rule issued by OSHA act in a way that is synergistic with this process rather than preempting it. Federal and state preemption on tobacco control issues has historically only served the interests of the tobacco industry. For example, without the federal preemption on cigarette advertising controls, many states and localities probably would have enacted much more effective measures to reduce promotion of tobacco to children than exist today It is imperative that in issuing its rule, OSHA be cognizant of ~ fact and create a rule that: (1) establishes a minimum standard rather than a preemptive federal standard, and (2) can work in tandem with local legislation and local enforcement mechanisms. Such a process will stimulate more public interest and public involvement in the process and lead to more effective and more cost-effective measures to reduce the exposure of workers to environmental tobacco smoke in the workplace. Conclusion There is very strong evidence that passive smoking causes or aggravates heart disease, particularly when compared with the level and quah'ty of evidence available for implicating o~ f~r rea~peo~ expose~ to ~ ~d~ ~ ~ T102322.239

both humans and ~ documenting short term effects of envimum~tal tobacco smoke on the cardiovascular system and showing that these changes~ can be mazfi.'pulated experimentally. Finally, there is good evidence implicating some of the specific physicuchemical agents in the tobacco smoke to explain ~mme of the~e effects. Using different methodologies, several investigators have estimated the population burde~ assodated with passive smoking and heart disease, yieldi~ estimates of 30,000- 60,000 deaths annually in the United States, with about three times as many non-fatal cardiovascular events. This is a tremendous public health impact and one that warrants strong action by OSHA to protect workers and the general public. Workers with existing heart disease are particularly at risk. In addition the toxins in ETS can act synergistically with toxins from other sources, and this possfoility should be considered in writing any rule. The simplest and most cost effective control measure is to mandate smoke-free work places, as is in the proposed OSHA rule. T102322240

References 1. US "Envirc~mmental Protection Agency. Respiratory health effecu of passive .smoking: Lung cancer and other disorders. US Environmental Protection A~_~,ency, 1992. USEPA Document .No. EPA/600/6-90/006E ." Wells AJ. An estimate of adult mortality in the United States from passive smoking. Env/ron. In~ 1988;14:249-265. Glantz SA, ParmJey WW. Passive smoking and heart disease: Epidemiology, physiology, and biochemistry. C'ucuJm/on 1991;83:1-12. National Institute of Occupational Safety and Health. Environmental tobacco smoke in the workplace: Lung cancer and other health effects (~ intelligence bulletin 54). National Institute for Occupational Safety and Health, 1991. DHHS(NIOSH) 91-108. Steenland IC Passive smoking and the risk of heart disease. JAMA 1992;267:94-99. Leone A. Cardiovascular damage from smoking: A fact or belief?. Int. J. Can//oL 1993;38:113-117. Wells AJ. Passive smoking as a cause of heart disease. J. Am. Col~ Card. 1994;24:546-554. Taylor AE, Johnson DC, Kazemi H. Environmental tobacco smoke and. cardiovascular disease: A position paper from the Council on Cardiopulmonary and Critical Care, American Heart Association. Circulation 1992;86:1-4. Gidding S, Morgan W, P-,:a'ry C, Isabel-Jones J, Bricker JT. Active and passive o tobacco exposure: A serous pediatric health problem. A staxement fxom the Committee o~ Ath~ an~ Hyperte~on in ~ ~ T102322241

10. 11. 12. 13. 14. 15. 16. ~A;(/n pr~). Wu-W'dliams A, Samet J. Envimnm~tal tobacco smog: Exposure-response Vander A.I, Sherman JH, LucianoI~Human Pi~'olo~: Tt~ Medmn~ms ~ Body Function. (5 ed.) New York: McCrrmv-Hill, 1990 Benowitz NL. Nicotine and coronary heart d/sease. Treru/s Card/cscr~ Med 1991;1:3B-~21. Zhu B-Q, Sun Y-P, Sievers RE, Isenberg WM, Glantz SA, Pm'mley WW. Increased exper/menml athemsclemsis in cholesterol-fed rabbits exposed to pass/ve smoke: Taking issue w/th study design and methods of mmlysis (reply to letter)./. Am. Co/~ Card/o~ 1993;22:17~2-17~3. US Department of Health and Human Services. The Hea/th Consequences oJ Smoking: Cardio~,ascular Disease A report of the 5'urgeon General. Public Health Serv/ce, Office on Smok/ng and Health, 1983. DHHS(PHS) 84-50204. US Department of Health and Human Services. The Health Consequences of Inmluntary Smoking.: A Ret~ort o.f the Surgeon ~enen~l. Office on Smoking and Health, Centers for Disease Control, Public Health Serv/ee, 1986. DHHS(CDC) 8/-8398. Moskow/tz W, Mosteller M, Schieken R, et al. Lipoprot¢in and csygen transport alterations in psssive smok/ng preadolescent children: The MCV Irwin study. Ckada~n 1990;81:~86-592. T102322242

17.. 18. 19. 20. 21. 24. 25. Environmental Protection Agency, 1990. USEPA Document No. Le, o~ A, Mori L, B~ F, Fabiano P, Filipp~ M. Indoor passiv~ ~: Its ef~ct on cardiac performance. Int. J. Card/oL 1991;33:247-252. Pomrehn P, Hollarbush J, Clarke W, Lauer R. Children,s HDL-choh The effects of tobacco, smoking, smokeless, and parential smoking. Circulation 1990;NEED CITATION. Moskowitz WB, Mosteller M, Hewitt ~ Eaves I.J, Nance WE, Schiek~n RM. Univatiate genetic analysis of oxygen transport regulation in children: The Medical College of Virginia twin study. Ped/a~. Rra. 1993;33:645-64S. Feldman J, Shenker IR, Etzel RA, et al. Passive smoking alters lipid profiles in adolescents. Pedhm/c.~ 1991;88:259-254. Gvozdjak J, Gvozdjako~ A, Kncharska, Bada V. The ¢t~ct of smoking on myocardial metabolism. ~ Meal 1987;10:47-53. Gvozdjakova A, Kucharska J, Gvozdjak J. Effect of smoking on the oxidative processes of cardiomyocytes. Cardiolo~ 1992;1992:81-84. Khaifen ES, Klochkov VA. Effect of passive smoking on physical tolerance of ischemia heart disease patients. Ter/lrkh 1987;59:112-115. Aronow W. Effect of passive smoking on angina pectoris. N End,/ ] Meal 1978;299:21-24. T102322243

Offtce of ~tal Health ~ A..ss~ssmea~ Can~,w.~ular Health Effects of ExOosure to Environmental Tobacco Smoke (Public Review Dra~). Ca~rnia Enviromental Protection Agen~ 1994. 27. McMurray RG, Hicks I.,L, Thompson DL. The effects of~passive infi~'~ation of cigarette smoke on exercise performance. Eur J Appl Phy~l 1985;54:196-200. 28. I.~mb D. Physiolog~ of Eren:ise: Responses and Adaptation.New York: MacMillan Pub~ Co., ~984 29. Pittilo RM, Mackie H, Rowles PM, Machine SJ, Woolf N. Et~cts of cigarette smoking" on the untrastructure of rat thoracic aorta and i~s ability to produce pmstacydin. Thromb. Haemo~-tas. 1982;48:173-176. 30. Davis J, Shelton L, Watanabe I, Arnold J. Passive smoking affects endothelium and platelets. Arch Intern Med 1989;149'.386-389. 3L Sinzinger H, Kefalides A. Passive smoking severely decreases platelet sensitivity to antiaggregatory pmstaglandines. Lancet 1982;2:392-393. 32. Burghubvr O, Punzengruber C, $inzingcr H, Haber P, $ilberbauer K. sensitivity to prostaoyclin in smokers and non-smokers. Che~-t 1986;90:34-38. 3"3. Sinz~er H, Virgolinl L Are passive smokers at greater risk of thrombosis? W/en Wochenschr ~989;20:694-698. 34. Steinberg D, Parthasarafl~y S, Carew TE, Khoo JC, Witztum JL. Beyond Cholesterol: Modifications of low-density lipoprotein that increase its athcrogcnic/ty. N. Eng£ f. Meal. ~989;320:915-924. 35. Martin JF, Bath PM, Burr ML. Influcmce of platel~t s/ze on outcome aftex n~y0.card/a] ~ Lancet 1991;338:1409-141L 4~ T102322244

37. 38. 39.' 40. 41. 42. artery disease. Y Lab C//n Meal 1985;105:479-483. Davis JW, Shelton L, F.~enberg I:~ I-li_gnite CE, Watanabe IS. Effect, s of tobacco and non-~obacco cigarette smoking on endothelium and platelets. C'lin P~I Ther 1985;37:529-533. Davis J, Shelton L, Hartman C, Eigenberg D, Ruttinger H. Sin. oking-induced changes in endoth¢lium and platelets are not affected by hydruxyethlmtosides. Br~ F.zp PathoI 1986;67:765-771. Davis JW, Shelton L, Eigenberg DA, Hignite CE. Lack of effect of aspirin on cigarette smoke-induced increase in circulating endothelial cells. Haemosta6s 1987;7:6669. Zhu B-Q, Sun Y-P, Sievers R, Isenberg R, Glamz SA, Parlmey WW. Passive smoking increases experimental athemsclerosis in cholesterol-fed rabbits. ~ Am Co//Cardiol 1993;21:225-232. Sun Y-p, Z.hu B-cl, SIevers RE, Glantz SA, Parmley WW. Metoprolol does not attenuate atherosclerosis in lipid-fed rabbits exposed to envimnmenfl tobacco smoke. C/~at/on 1994;89:2260-2265. Zhu B-q, Sun Y-p, Sievers RE, Glantz SA, Parmley WW, Wo~ CL Exposure to envhxmmcntal tobacco smoke increases myocardial infarct size in rats. C/nu/at/on 1994;889:1282-1290. T102322245

1992;70:341-347. 44. Benowitz NL, Fitzgerald GA, Wilson M. ~ Q. Nicotine effects on eicosanoid formation and hemostatic function: Comparison of transdermal nicotine and cigarette smoking. ]. Arr~ Coll. Cardiol. 1993;22:1159-1167." 45. Prerovsky I, Hladovec J. Suppression of the desquamating effect of smoking on the human endothelium by hydroxyethylrutosides. B/ood Ve.sse/s 1979;16:239-240. 46. Benditt E, Benditt J. Evidence for a monoclonal origin of human atherosclerotic plaques. Proc NatI Acad 5ci USA -1973;70:1753-1756. 47. Albert RE, Vanderlaan M, Burns FJ, Nishimu~ M. Effect of carcinogens on chicken atherosclerosis. Cancer Res 1977;37".2232-2235. 48. Revis N, Bull R, Laurie D, Schiller C. The effectiveness of chemical carcinogens to induce atherosclerosis in the white carneau pigeon. Tox/co/o~ 1984;32:215-227. 49. Penn A, Batastini G, Soloman J, Burns F, Albert R. Dose-dependent size increases of aortic lesions following chronic exposure to 7.12-Dimethybenz(a)anthracene. Cancer Res 1981;41:588-592. 50. Penn A, Garte S, Warren L, Nesta D, Mindich B. Transforming gene is human atherosclemtic plaque DNA. Proc Natl Acacl Sci USA 1986;83:7951-7955. 51. Majesky M, Yang H, Benditt E. Carcinogenesis and atherogenesis: Differences in mono~ygenase inducibility and bioactivation of benzo(a)pyrene in aortic and hepatic tissues of atherosclero.sis-susceptible versus resistant pigeons. Carcinogenesis 1983;4:647-652. T102322246

56. 57. 58. 59. 60. Natkmal Research Council Committee on Passive Smokit~ ~ Tobacco Smog- Mem,.a~g Ex~ and~ Health E~.Washingtc~ DC: National Academy Press, 1986 Ott W,/_andan L, Switz~ P. A time series model for cignxette smoking activity patterm: Model validation for carbon monoxide and respirable particles in a chamber and an automobile, d. Exposure Anal Environ. EpidemioL 1992;2(Suppl 2):175-200. Wu JM. Increased experimental atherosclerosis in cholesteml-_fed rabbits exposed to passive smoke: Taking issue with study design and methods of azmlysis (letter)..L. Am. Coll. CwdioL 1993;22:1751-1752. Penn A, Snyder CA. Inhalation of sidestream ciagrette smoke accelerates development of artiosclerotic plaques. Cimulmion 1993;88(Part 1):1820-1825. Penn A, C"hen LC, Snyder CA. Environmental levels of sidestream cigarette smoke promote arterioscaemtic plaque development (abstract). ,Sot/ely of To.t/co/o~ and Experimental Biology. Anaheim, CA:, 1994:. Penn A, C'm'rie J, Snyder CA. Inhalation of carbon monoxide does not accelerate arteriosclerosis in cockerels. Ew:. ]. Pharmacol. 1992;228:155-164. Hc~vard G, Szklo M, Evans G, Tell G, Eckfeldt J, Heiss G. Passive smoking and carotid artery wall thickness: The ARIC study (abstract). C/mu/at/on 1992;85:862. • Church DF, Pryor WA. Free-radical chemistry of cigarette smoke and its toxicological implications. Envirn. Health Persp. 1985;64:111-126. Ferrari R, Ceconi C, C'u~lio S, et al. Oxygen free radicals and myocardial damage: Protective role ~ thiol-conmining agents.Am. J. Meal. 1991;91 (suppl 3C):95S-105S. T102322247

6L 2. 63. 4t 5D 66. 68. Przykiea~ K. Nicmi~ ~tes postisc~emic contmeu'le ~dy~mtion of'~ myocardium in the canine model: Possible role fo free radicals. ~ 1994;89:1272-128L van ffaarswad H, Kuyl JM, Alberts DW. Exposure of rats to low concentration of cigarette smoke increases myocardial sensitivity to ischaemia/re~rfusiom Ba,~ic Res. Cardio/~ 1992;87:393-399. van Jam'sveld H, Kuyl JM, ALberts DW. Antio~idant vitamin supplementation of smoke-exposed rats partially protects against myocardial ischaemic/reperfusion injury. Fee Rad. Res. Comras. 1992;17:263-269. McC~ker K, Hoidal J. Selective increase of antioxidant enzyme activity in the alveolar macrophages from cigarette smokers and smoke-erposed hamsters.Am. Re~ Respb: D/~ 1990;141:678-682. Anderson R, Theron AJ, Richards GA, Myer MS, van Rensburg AJ. Passive smoking by h~ sensitize~s circulating neutrophils. Am. Rev. Respb:. D/s. 1991;144".570-574. Prentice RC, Carroll R, Scanlon PJ, Thomas JXJ. Recent exposure to cigarette smoke increases myocardial infarct size (abstract). Y. Am. Coll. Cardiol. 1989;13:124A. Otfil]en JE, Rossen JD, Oskarsson HI, Minor RLJ, Lopez AG, Winniford MD. Acute effect of cigarette smoking on the coronary circulation: Constriction of epicardial and resistance vessels../. Am. Co/t Cardiol. 1993;22:642-647. Fenton RA, Dobson JGJ. NicolJne increases heart adenosine release, oxygen consumption, and contracl~ty. Am..l. Physiol. (Heart) 1985;249:H463-H469. TI02322248

Coll. Cardiol. 1993;22:1881-1886. 70. Garlsnd C, Barrett-Connor E, Suarez L, Criqui M, Wingard D. Effects of.passive smoking on ischemic heert disease mortality of nonsmokers. Am ] Ep'u~emiol 1985;121:645-650. 71. Hole DJ, ~ CR, Chopra C, Hawthorne VM. Passive smoking and cardioresp/ratory health in a general p0opulation in the west of Scotland. Br Med ~ 1989;299:423-427. 72. Humble C, Croft J, Gerber A, M. C, Haines C, Tymler H. Passive smoking and twenty year cardiovascular disease mortality among nonsmoking wives in Evans County, Georgia. Am ] Pub Health 1990;80-.599-60L 73. Jackson RT. The Auckland Heart Study. University of Auckland (New 7_,ealand), 1989. 74. Sandier DP, Comstock GW, Helsing KJ, Shore DL. Deaths from all causes in in nonsmokers who lived with smokers. Am../. Pub. Health 1989;79:163-7. 75. Helsing K, Sandier D, G. C, Chee E. Heart disease mortality in nonsmokers living with smokers. Am ] Epiderniol 1988;127:915-922. 76. Butler TI_ The relationship of passi~ smoking to mrious health outcomes among Seventh Day A~,dentl.sts in California. Ur~-versity of California, Los Angeles, 1988. 77. Butler T. The relationship of passive smoking to various health outcomes among Seventh-Day Adventists in California (abstract). Seventh World Co~ on Tobacco and ~. P,-~-'th, ~a:, 1990:. 316. T102322249

80. 8L 82. 83. 85. 85. 87. ~ T. Passive smoking. NZ Meal. J. 1990;103-.54. Sv~nds~ KH, Kall~ LH, Mania MJ, Ocl~n.~ ~ ~ o~ passive smoking in the Multipl~ Risk Factor I~te.rvention Trial. Am J Ep/dem/ol 1987;126:783-795. Hunt $C, Martin MJ, W~ RR. Passive smoking by nonsmoking wiws is assodated -with an increased/nddence of heart disease. Amedcan Pubfic .Health "" A.~oc/at/on. Las Vegas, NV:, 1986:. Palmer J, Rosenberg L, Shapiro S. Passiv~ smoking and myocardial infarction. Epidemiol New~letter 1988;43:29. He Y, Li' L-S, Wan Z-H, Li L-S, Zheng X-L, Gru L-L Womeff~ passive smoking and coronary heart disease. Chung-Hua-Yu-Fang-l-Hsueh-Tsa.Chin 1989;23:19-22. He Y, Lain TH, Li TS, et al. Passive smoking at v~rk as a risk factor for coronary heart disease in Chinese women who have n~ver smol~d. Br. Meal: 1994;308".380-384. Dobson AJ, Alexander HM, Heller RF, Lloyd DM. Passive. smoking and the risk of heart attack or coronary death. Meal. L Aust~. 1991;154:793-797. Le~ P. Deaths from lung cancer and ischaemic heart disease due to passive smoking in N6w Zealand (letter). NZ Meal J 1989;102:448. LaVecchia C, D~vanzo B, Franzosi MG, Tognoni G. Passive smoking and the risk of acute myocardial infarction (letter). Lancet 1993;341".505-506. Bcro I.A, Glantz SA. Tobacco industry response to a risk assessmeat of environmental tobacco smol~. Tobacco Control 1993;2:103-113. Be~ I..A, Gha~ SA, Reanie D. Publication bias and public health policy on 47 T102322250

89..BeroLA, GaroraithA, ~D. Spomored~oaeavimmma~tobecco smoke. JAMA 1994;271:612-617. 90. Remnie D. Smoke and letters~ JAMA 1993;270:1742-1743. 9L Rennie D. Smoke and letters (letter). JAMA 1994;271:1575. 92. US Department of Health and Httman Services. The Health Benefits of Srdoking Cessatio~ A Report of the Surgeon General. Office on Smoking and Health, Centers for Disease Control Public Health Service, 1990. DHHS(CDC) 90-8416. 93. --,. Self-serving surveys: The 30 percent myth. Consumer Reports 1994 May, 1994:320. 94. Glantz S, Smith LRA. Effect of smoke-f~ee restaurant ordiances on restaurant sales. Am. J. Pub. Health 1994;84:1081-1085. 95. Siegel M. Involuntary smoking in the restaurant workplace: A review of employee exposure and health effects, lAMA 1993;270:490-493. 96. Pierce JP, Shanks TG, Pertschuk M, et al. Do smoking oxdinances protect non-smokers from environmental tobacco smoke at work? Tobacco Control 1994;3:15-20. 97. Pierce JP, Evans N, Farkas AJ, et aL Tobacco Use in Cali.~'nia: An Evaluation of the Tobacco Control Program, 1989.1993. University of California, San Diego, 1994. TI02322251