Tobacco Documents Online

JACC Vol. 21l No. I January 1993=5-32 ar, J J j rw' Cl,a css- «an. EXPERIMENTAL STUDIES Passive Smoking Increases Experimental Atherosclerosis in Cholesterol-Fed Rabbits BO-QING ZHU, MD, YI-PING SUN,, MD, RICHARD E. SIEVERS, BS, WILLIAM M. ISENBERG, PHD, STAI*TTON. A. GLANTZ, PxD, FACC, WILLIAIvf W. PARMLEY, MD, FACC San Francisco, California Objectives. We evaluated'the influence of passive smoking on experimental atherosclerosis in cholesterol-fed rabbits. Background. Exposure to environmental tobacco smoke (ETS)) has been ~ epidemiologically linked' to death from ischernic heart disease in nonsmokers. Methods. New Zealand male rabbits were randomly divided into tltree groups after 2 weeks of a 0.3% cholesterol i diet. Sixteen rabbiu were exposed to a high and 16 rabbits Wa low dose of ETS; 32 rabbits located' in another room served as an unexposed control group. After 10 weeks of ETS exposure, all rabbits were killed, and the percent of aortic and pttlmonaryarteryendotheiial surfaces covered by lipid lesions was measured by, staining and planimetry.. Results. Average air nicotine, carbon monoxide and total particulate concentrations were 1,040 Ecglm3', 60.2 ppm and 32.8 mgim3 for the high dose ETS group, 30 pglm3,18.8 ppm and 4.0' mglm3 for the low dose ETS group and <1 µgimi, 3.1 ppm an&0.13 mg/m3 for the control group. The percent atherosclerotic Environmental tobacco smoke (ETS) is the term used to describe tobacco combustion products inhale& by nonsmok- ers in the proximity, of burning tobacco. More than 4,000 ,;onstituents have beem identified in cigarette smoke. Most, ~ c,;u~ure to ETS is from sidestream smoke emitted from the blltning tip of the cigarette. Sidestream smoke is hazardous because it contains high, concentrations of ammonia, ben- zene,, nicotine, carbon monoxide and many other carcino- gens and irritants (1-3). Passive smoking - involves breathing both sidestrearn smoke that goes directNy into the aib from the burning tobacco products and~ mainstream smoke after it has been exhaled ~ by smokers. Sidestream smoke has higher concentrations of nox- F,-om the Cardiovascular Divisioo. Department of Medicine and the Cudio- vascular Re:search Ihstitute, Utuversity of Californiz San Fnncisen, Caldontii This study was supponed in part by the Tobaeeo Reimed Disease Rcsnrzh ~ Progmii (Gran[~ 1 RT 145), University of Caldornia and the George Smith Fund. All editonal'decisions for this arucle, including selection of referees, were tnade by a Guest Editor. This policy appl)es to aU utieles with authors from the Uaiversity of California, San Francisco, Maausenpt received Much31, 1992: revised manuscript received July 7, 1992„accepted July 15, 1992. Address for corresuortdence: William W: Parmky, MD, 1186 MotUtt Hosprtal. Uui.+ersttyof Califomtn, San Francisco. San Francisco, Califomia 94143-0124. involvement of the aorta and pulmonarv artery increased signi5- nntly with ETS exposure (for the aorta, 30'± 1996 [mean y SDj for the control group, 36 ~: 14% for the low dose ETS group and, 52 = 21% for the high dose ETS group, p< 0:D01; for the pulmonary artery, 22 = 15% for the control group+ 29 i 25% for the low dose ETS group, an& 45 _ 1296 for the high dose ETS group, p< 0.001): Bieeding tiine was significanUyshorter in the two ETS groups than in the control group (86 = 17 vs. 68 = 15, 68 ~: 18 s, p < 0.001). There were no: signi5eant differences in serum trigiycerides,,chole-sterol and'high density lipoprotein rbo- lesterol at the end of the study. Conclusiont. Environmental tobacco smoke affectU platelet function and increases aortic and pulmonary artery atberosclero- sis. This increase of atherosclerosis was independent of chaages in serum lipids and exhibited a dose-response relation. These results are consistent with data from epidemiologic studies demonstrating that ETS increases the risk of death due to heart disease. (J' A m Coll Cardio! l993;21:22,5-32) ious compounds than does mainstream smoke. lt has been estimated that approximately 50 million: nonsmoking adults over the age of 35 years are regularly exposed to environmental tobacco smoke. Addiuonally, 50% of all chilliren live in fami- lies with one or more smokers (4): The effects of passive smoking on health have been reported to include short-term~ effects, such as exacerbation of asthma and angina, as well as long-tetm~effects, such as increased~ risk of lung cancer, respi- ratory tract infection and atherosclerosis (1-7). Environmental tobacco smoke adversely affects platelet function and damages arteriali endothelium, and depresses cellular respiration at the level of mitochondria (4',5). People exposed toit have significantly thicker anenal walls than~do unexposed nonsmokers, and wall thickness is increased with increasing exposure (8). Passive smokers also have signifi- cantly depressed high density lipoprotein (HDL),cholesterol levels and significantly elevated ratios of total cholesterol to HDL cholesterol levels. (9). The materials in ETS may thus accelerate the develop- ment of atherosclerotic plaque. Previous experimental studies, however, showed' that exposure to smoke from only 1 cigarette/day for 11 to 13 months failed to quantita- tively affect atherosclerosis or serum lipids (S0). We de- signed the present study toi further evaluate the influence 0735- I t)97/93156.00 3606 C 1993 by the American Coneae of Cardiology

2~5 ZHU ET AL. PASStVESMOR]NG INCREASES ATHEROSCLEROSIS of passive smoking on atherosclerosis in cholesterol-fed rabbits. Methods Protocol. Sixty-four New Zealand male rabbits (2.0 to 2.6 kg) were randomly separated into three groups and fed a high cholesterol! diet for 12 weeks. The cholesterol diet. (Ziegler Bros., Inc.) contained 3% soybean oil and 0.3% cholesterol by weight. The rabbits were housed in separate cages in well mize& exposure chambers (BioClean, Duo. Flo, model H 5500;, Lab Products Inc.), 1.92 m x 1.92 m~ x 0.97 m (3.58 m), that accommodated eight rabbits in each group. After 2 weeks on the diet, 16 rabbits, 8 at a time, were exposed to a high dose of sidestream smoke (high ETS group) from Marlboro filter cigarettes (4 cigarettes every, 151 mim for 6 h/day, 5 days/week) using a smoking machine (Heinr. Borgwald GMBH' RM I/G, D-2000 Hamburg, Ger- many) for 10 weeks from week 2 to week 12. Another 166 rabbits. 8 ao a time, were given a low dose of smoke Qow ETS group) from the same smoking machine through 20.5 feet ofl 10-mm inside diameter plastic tube attached to the mainstream port on the smoking machine. The smoke cooled and the large particles settled out in this tube, making the exposure level of the low ETS group similar to that of smoke spread by the ventilation system of a building from an~ area where smoke was permitted to nonsmoking areas of the same building. Thirty-two rabbits, 16 at a time, located in the same type of exposure chamber in another room~but with no smoking machine, served as a control group eating the same diet for 12 weeks, Three fans in the exposure chambers were adjusted to ensure good mixing, using the measurement devices discussed later. At the end of the 6-h exposure period', the exhaust fan on the Biocleam unit was turned' on and rapidly lowered the level of ETS pollution in the exposure chamber to background levels corresponding to those of the control animals until next day when the$ioclean unit was turned off and~ the smoking machine was turned on again. Monitoring smolieezposure insid'e the ehambers. We mea- sured several constituents of ETS in the three exposure chambers: carbon monoxide (CO), total particulates„respi- rable suspended patticulatcs and nicotine. To measure average carbom monoxide concentrations during the 6-h exposure period, we used a model L15 CO Personal Exposure System (Langan Products), every other week for the three groups. We obtained an average daily valtie taken from 2,520 samples during the exposure period (3 h of ETS, I h break, 3 h of ETS) (Fig. 1). To measure total particulate concentrations, we used' a Miniram PDM-3 Optical Scattering Particle Monitor (MIE, Inc.), monitoring particulate concentration every 1D~s, and computed average total particulate concentrations during the exposure period (Fig. 2): We obtained1hese data every other week for all three groups. We also used a Piezobalance JACC voll :1. No. I Ianuary,1993125-32 Carbon Monoxltl. (ppm) roa 95 90 e5 6o4 Hgh -B7 75 7 o 65- 6a 55 sa- 45 .0-1 35 30 25 20 ,5 ro 5 Exposur• p.rlotl. Figure 1. Representative carbon monoxide (CO) concentrations during a 24h period. During the period of exposure to environmen- tal tobacco smoke (ETS) (3 h~of ETS, a 1-h break, 3 h,of ETS); the average CO value from 2,520 samples is 53 ppm. Respirable Aerosol Mass Monitor (model 3500 Thermo- System) to measure respirable suspended particulates (11l) on 4 different days, about, 10 sampieslday; to calibrate the Iviiniram. The Piezobalance was factory, calibrated before the study. The Piezobalance measures the smaller respirable,sus- pended particulates, whereas the Miniram measures total particulates. To determine the relation between particulate concentrations measuried;by the Miniram and Piezobalance, we measured, average particulate concenuation values (37 values, each an average of~ 3 measurements) at different levels of ETS using these two instruments simultaneously. Figure 3 shows that there was a strong linear relation between average particulate concentrations measured by the Piezobalance and the Miniram, with the Piezobalance read- ing about 36~'0 of that obtained by, the Miniram. This relation is sinular to that found in a previous study (12) in which the Miniram and the Piezobalance were compared in an envi- ronmental chamber measuring ETS over a range of concen- trations. In addition, we monitored air nicotine levcls by using a passive diffusion monitor~ (13)ithat was located in the middle of the exposure chamber during the 6-h exposure period, every other week for all three groups. Hematologic and biochemical analysis. Bleeding time. cii'- culating platelet aggregates, platelet count. hematoerit. he- moglobin, totai'serum cholesterol, triglycendes. high density lipoprotein (HDL) cholesterol and serum cotinine were measured at the beginning of the study (before the rabbit3 started the high cholesterol diet) and at 6-week intervils (that is, after 4 and' 10 weeks). The concentration d aodnine was determined by gas chromatography with nitrogca- phosphorus detection (14): This method has been modified for simultaneous extraction of cotinine and detcratiaaboo using capillary gas chromatography (15). Bleeding time was determined after l'-min warming of the rabbit's ear in a normal saline bath (37°C)1 A smaA stand9rd

JACC Vol. a. No. I January 19912:3-32 n Figure 2. Total suspended particulate eoncentrations measured by the Miniram every, 10 s during a representa- tive peno&of exposure to environmental tobacco smoke (ETS) (3 h of ETS. a I-h break, 3 h of ETS). Average total suspended particulate concentration ~ during the exposure penod' was 23.7 mgim'. The peaks occur while the ciga- rettes are actually being smoked. The large drop corre- sponds to ithe 1-h middaybreaky prick was made in the ear, avoiding macroscopic vessels. The time fromithe initial bleeding to cessation of bleeding was recorded as the bleeding time. A platelet count ratio method (16) was used for, quantita- tive determination of circulating platelet aggregates. One mmoUliter adenosine diphosphate was added to a citrated venous blood' sample before stirring, The sample was di- vided into two tubes, one containing ethylenediaminetet- raacetic acid (EDTA)/fotmalin~solution and the other EDTA only. Platelet-rich plasma was collected after centrifugation. Platelets im both samples were counted, using standard techniques (Sequoia-Turner Corporation„ Operator Refer- ence manual, Cell-Dyn 900 Hematology Analyzer). The platelet aggregate ratio was calculated' from the platelet count in the two solutions. The higher the ratio, the fewer the platelet aggregates. Figure 3: Relation~between average particulate concenuations mea- sured by the Piezobalance (PZB) and the Miniram. Because of the ezcellent llnear relation~ one can measure respirable suspended particulates by taking 36Wof the readings obtained with the Mini- ram. . Ar.r.yoportlea aone.nrnt/aWnIpnn,tJ) blWftkaM (1IJ ZHU ET AL. 227 PASSIVE SMOKING INCREASES ATHEROSCLEROSIS Tfmr Total serum cholesterol and trigl,Ncende levels were de- termined by automated enzymatic methods (Coulter DART cholesterol reagent using the DACOS and DACOS XL analyzers), and HDL cholesteroliconcentrations were mea- sured after precipitation of other, lipoprotein classes with dextran and magnesium ions (HDL cholesterol precipitant (CacNo 236141)„Ciba Corning Diagnostics Corp.). ' The blood samples were drawn in the morning (Tuesday to Friday) after 12 h,of fasting and before ETS exposure. The samples for plasma cotinine analysis also were taken,imthe morning before exposure (17 h after the last ETS exposure). Morphologic studies. At: week 1?, after 10'w•eeks of exposure to ETS (or control conditions), all rabbits were killed. After intravenous administration of pentobarbital, 130 mg/kg body weight, the aorta was removed from its origin (2 cm dista1toithe aortic valve) down to the bifurcation of the internal' iliac arteries; the pulmonary artery was isolated from its beginning at the pulmonary vaJ4e to just above the bifurcation. The vessels were opened by linear vertical incision, fixed in a 104io formalin solution for 24 h, stained with Sudan IV', then photographedl Finally, plani- metric measurement of lipid lesions was performed quanti- tatively by estimating the total stained' regions in photo- graphs of each artery with a planimeter. The measurements were performed in blinded'fashion and imduplicate. Statistical analysis. The text and tables list data as the mean value = SD; the figures summarize data as the mean value = SEM. Data were analyzed by linear regrzssion.. using ETS dose as the independent'variable. Multiple linear regression~was also used with aortic and pulmonary artery lesions as the dependent variables, including choksterol levels as well as exposure to smoke in the regtssion equation to account for the possible effects of dtfferrnt serum ~ cholesterol levels on the extent of lesions. Analysts of variance (ANOVA) was used to compare observations among the three experimental' groups. Data were analyzed before and after exposure, as wclllas in terms of changes in

22$ Z.NU ET AL. , PASSIVE SMOKING INCREASES ATHEROSCLEROSIS JACC Voll'-1. No. I January 1993:225-32 Table 1l Average Air Nicotine, Carbon Monoxide and Particulate Concentrations in the Control and Environmental Tobacco, Smoke Chambers Group Air Nicotina (pglm') Air CO IPpm)! Total P.rticulates' (mgrm3) Resptrahle Particutatest, (mg!m3) Control <1 In = 1!1 3.1 s 1.9 (n = 2) 0:13 _ 0!04 In = 10)', 0.07 t 0.06 (n = 10), LowETS 30=31n=4), 18.8c2.2(n=5) 4.03z0149(n=3) 1.2=0.7In=8) High ETS 1.040 _ 302 (n = 4) 60.2_ 14.3(n=5) 32:8z6.9(n=7) 13.8=3.5(n=6) 'By Miniram: *By Piezobalanee. Values are expressed as mean value _ SD. ni= the number of samples. For nicotinc, carbon monoxide [CO] and'total paruculates, each of the n samples represents the average value observed during the exposure period (3'h of environmental tobacco smoke [ETSl, lib break. 3 h o6 ETS)',on I' day. For examplc„the n= 2 values for Air CO in the controL group represenraverage values recorded during 7 h on 2 diffcrcnt days. For respirable particulates, the sample size represents the actual number of simple samples taken while the smoke Ievel's were at steady state. the measured variables before and after the 10-week expo- sure perio& using paired r tests. We did not combine all data into a single two-factor analysis of variance (with time [before or after exposure) as one factor and ETS group. [control, low ETS, high ETS]! as the second~ factor): we believed that with~ such an approach~ the presence of a control group (with no exposure) at both times would generally lead to a significant interaction between time and exposure group that would make the results of'tests on the maini effects difficult to~ interpret. Data were processed by using Minitab Versions 7.2 and 8.2. A p value < 0.05 was taken as statistically significant. Results Weight gain. There was a similar initial body weight and ~ttl s?qu nt w,~ight gain in all'rthree groups of'irabbits througli- olt(I_7~,voak period. The average body weight before (week 2) and after 10 weeks of ETS exposure (week 12) was. 2.7 = 0.3 and 3.6 = 0.3 kg, respectively. There was no significant difference in weight of the rabbits as assessed by Ai`+t)V,\ before (p = 0.344) or after (p = 0.306) the 12-week exp,;,iutcnt;al period. Similarly, ANOVA showed no signi6 cant differences in weight gain among the three exposure groups (0.87 = 0.29 kg for the control group„0.88 - 0.37 kg for the low ETS group and 0;91 ± 0.36 kg for the high ETS group; p= 0.923) or in food intake among the three groups, either before (p = 0.398) or after (p = 0.43 1) exposure to ETS. The average food intake before and after ETS expo• sure was 178 - 46 and 164 ± 58 glday, respectively. The similarities in eating and weigttt gain across time and the Tabl'e 2. Effects of Environmental Tobacco Smoke on Serum Lipids in,Choiesterol'.Fed Rabbits Cboltsterol (mg/dl) Group Before After Conuol la ; 32) 671 = 278 1r09 - 483 Low ETS (n - 16) 480 : 279' 1.154 = 395 HighETS(n = 16Y 531 s 246 1.260 =532 different exposure groups indicate that any differences ob- served in the exposure groups were not due to dietaryy differences. There were no deaths during the 12-week study. Smoke exposure inside the chamber. The average air nicotine, carbom monoxide (CO) and total particulate con- centrations during the 6-h exposure period are listed in Table 1. There were lrirge differences in air nicotine, CO and particle concentrations between the groups with a high or low leveli of ETS exposure and the controll group and between the high and low ETS groups during the period of exposure. Altetrations in lipids. After rabbits were fed a high lipid! diet, the serum cholesterol increased considerably in all' animals during the 12-week period. The serum lipid levels for the three groups (Table 2) show a similar increase in total serum cholesteroll Total cholesterol may have been sligtitly (p = 0.051) higher in the control group than in the two ETS groups before the 10-week exposure period. There was no significant difference (p > 018) among the three groups at the end of the experiment. There were no significant differences (p > 0.3) in triglycerides and HDL cholesterol among the two ETS' groups and the control' group either before on after the 10-week exposure period. There also were no significant differences (p > 0.4) in the area under the cholesterol time curve (cholesterol-weeks: 11,632' = 3,479'vs. 9,831 -- 3:048 and 10,349 t 3,182 mg/dl-wk), change (12-week value minus 2-week value) in cholesterol (538 = 463 vs. 674 - 419 and 729 - 627, 674 ± 419 mg/dl)„change in triglycerides (13 - 84 vs. -63 ± 372 and~ 22 = 91 mg/dl) and change in HDL jU Q cholesterol (15 = 29 vs. 7= 22 and 16 :t 25 mg/dl), ~ ~ Tnglycrndes Itog/dt) HDL Cbotesteml l(mg dl) ~ 2V , Before ARer Before After N 91 : 72 78 s 51 40 = 16 55 :'7 102 t 93 165 _ 349 36 s 13 43 : 25 119 = 93 98 : Ill 37 _ 15 50i- 21 Vklues are expressed as mean value = SD. There were no significanudiffertnees(p > 03)'.mons the three troups except for toul cholesterol beforc exposure to eovironmental tobacco smoke (ETS). Values in the control group were higher than values in the other two g'rwrps (p = 0.05). After = 12 weeks on lipid diet and l0'weeks of smoke exposure; Before = 2 weeks on Gpid'diet and before smoke exposure; HDL - high density, lipoprottin. W ~ j

JACC Voll 21. 1<a, l January 1993:'~t-32 , RSP Cone.ntnt/on,Dy. PZB (m01m') Figure 4. Relation between log-plasma cotinine levels and log- respirable suspended particulates (RSP) concentrations measured by Piezobalance(PZB); ETS-H and ETS-L = groups witha high or low level, respectively, of exposure to environmental tobacco smoke. respectively, among the control'group and the low and high, ETS groups. Cotinine levels in plasma. The plasma: cotinine levels at the 6th week of ETS exposure in the control and the low an& high ETS groups were <1.0, 6.0 t 4.3 and 1'S:6 = 12.3 ngJml, respectively. These cotinine levels are based on blood' samples drawn in the morning before that day's exposure to ETS. Given the 20-h~ half-life of cotinine in the blood, the steady state cotinine levels at the end of the daily exposure period would be approximately <1, 12 and 31.2 nglml,, respecuvel,v; for the control and low and high~ ETS groups. There was a linear relation betweendog cotinine levels in plasma and log average respirable suspended particulate concentrations measured by Piezobalance (r = 0!84, p < 0.001) (Fig. 4). Morpbologic studies. Figure 5 shows the percentage of' total aortic and pulmonary artery surface area covered by Figure 5. Percent of aortic an& pulmonary artery, surface areas covered' by atherosclerotic lesions for each group, There is a significant (p < 0.001) dose-response relation for both vessels. Error bars are SESt. Abbreviations as in Figure 4.  Conua  Ers-t  ETS•H Aorta iu4naosry Art.ry ZHU ET AL. 229 PASSIVE SMOKING INCREASES ATHEROSCLEROSIS lipid lesions in the three expenmental groups. There was a significann (p < 0.001) dose-response relation for, the extent of lipidilesions for both the aorta and the pulmonary artery as a functioniof'respirable suspended particulate concentration measured by Piezobalance. Although the intercepts of the dose-response relations for the two arteries are significantly different (31.3 - 2.7% for the aorta vs. 23.0 - 2.6% for the pulmonary artery, p < 0.05), the slopes are not (1.62 = 0.41%/t`ng%mt' vs. 1.69' - 0.39%imgim', p > 0.5). These results indicate that, although the baseline levels of lipid deposits in these two~ artenes are different, the effects of exposure to ETS on~the two anenes are similar in terms of increased lipid deposits. There were also positive correla- tions (r = 0.5; p< 0.001) between the percent of lipid lesions in both arteries and the average CO level5. As with the relation between lesions and particulate concentration. the aorta initiallN ha& more lipid deposits than did the pulmonary artery, but both vessels showed similar increases. (-0.5%/ppm) in lipid deposits with ETS exposure as,CO was increased. Because particulate and CO levels are highly correlated, we cannot say whether either or both (or other) elements of the ETS are responsible for, the dose-dependent increase in lipid deposits we observed. We can conclude unequivocally that there were significant (p < 0.0010idose- dependent increases in lipid~ deposits on both vessels with increasing ETS exposure. Platelet fnnction. Data on bleeding time. platelet aggre- gate ratio and platelet count are shown in Table 3. Bleeding times at week 12 in the low an& high ETS groups were significantly shorter than those in.the control group (68 = 15.. 68 = 1,8 vs., 86 = 17 s. respectively, p < 0.001). This result demonstrates that there were large (20%) changes in bleed- ing time at low levels of exposure to ETS and that further increases in~ezposure did not produce anadditional effect,. The platelet aggregate ratio at week 12 in the high ETS group may have been lower than the control level (79.4- 10.7 vs. 88.0 :t 12.2%, p = 0:07 by paired r test), reflecting an increase in platelet aggregates in the high ETS group: The platelet counts were modestly decreased to a similar extent in all three groups (Table 3). The changes in platelet count before and afterexposure were -36 = 97, -84 = 131 and -94 = 95 (p = 0:151 by ANOVA), respectively. These data show effects on platelet function at low levels of ETS that do not increase with further increases im dose. This result suggests that platelets are sensitive to~ low, levels of ETS. after which the effect is saturated'. Discussion Active smoking has consistently been, identified as a major risk factor for ischemic heart disease. Exposure to environmental tobacco smoke (ETS), as passive smoking, has now been linked to heart disease in nonsmokers (4.6.17- 19). Epid'emiologic studies conducted in a variety of loca- tions reflect about a 30~7o increase in risk of death from ischemic hean disease or myocardial infarction among non- ®

230 ZHl; : ET AL. , PASSIVE SMOKING INCREASES ATHEROSCLEROSIS _Table 3. Effects of Environrnental Tobacco Smoke on Platelet Function in Cholesterol.Fed' Rabbits JACC Vol. 21.,No, I January,1993125-32 Blteding Time (s) Platelet Aggregauond%) Platelet Caunt (1U'1. Group Before After Before After Before After Control In,= 32) 78 t 23 86 _. 17 84.1 _ 14.6 80:9 - t3.7 295 - 89 2.56 : 89 Low ETS In = 16)~ 73 t 26 68 t IS 83.9 _ 11.8 8Z6 _ 14.3 352 = 130 268 _ 95. High ETS in = 161 77 , 2 19 68 = 18 ' 87.9 t 12.3 79.4 : 10.7' 372 = 140 293 _ 76 •p < 0:01 compared with values in the controfgroupl ip = 0:07~ compared with values in the high ETS group before exposure. Values are expressed as mean value _ SD. Abbreviations as in Table 2. smokers living with smokers (4-6,18). The larger studies also demonstrate a significant dose-response effect, with greater exposure to ETS associated with a greater risk of death from heart disease. Our present study shows that passive smoking signifi- cantly increases aortic andipulmonary artery atherosclerosis in cholesterol-fed rabbits in a dose-dependent manner. There was a strong positive correlation between the percent of atherosclerotic lesions and the average CO or particulate concentrations„with the lipid deposits in arteries in~the high dose group nearly doubling in just 10 weeks. These results are consistent with epidemioiogic studies demonstrating that ETS increases the risk of death from heart disease. Passive smoking and atherosclerosis. Smoking has long been recognized as one of the major risk factors for adult ~ coronary heart disease, peripheral arterial disease, abdomi- nal aortic aneurysm and stroke. Clinical investigations in& cated that the proportion of intimal' surface involved with atherosclerotic lesions in both the aorta an&the right coro- nary artery was positively associated with serum very low density lipoprotein and low density :ipoprotein cholesterol and was negatively associatedvith serum HDL cholesterol. The serum thiocyanate concentration, a marker for smoking, was strongly associated with the prevalence of atheroscle- rotic lesions, particularly in the abdominal aorta (20). Popu- lation studies of passive smokers revealed that passive smokers had significantly thicker carotid arterial walls than those of persons who had never smoked passively or ac- tively (8). Our results are consistent with what one would expect from these clinical studies. However, we observed much larger effects of ETS than would be expected from a simple dose-based extrapolation from high doses experienced by smokers to tbe lower doses of smoke experienced! by nonsmokers. Our results suggestt that nonsmokers may be more sensitive to the toxins in ETS than smokers are, perhaps because smokers have somehow adapted to the chronic poisoning associated with active smoking. It is also probable that some of the biochemical systems involved are very sensitive to ETS but saturate at low doses. Passive smoking and serum lipids. Epidemiologic studies : have suggested that there is a dose-response relation be- tween the number of cigarettes smokedlday and increasing levels of plasma cholesterol'(2T). The HDL cholesterol level was lower in childien exposed to ETS; the HDL2 subffraction was reduced in boys, whereas the HDL3 subfraction was reduced in girls. As a result, exposure of children to ETS may increase the risk of premature coronary heart disease (22). Nonsmoking adolescents with two smoking parents had' significantly higher plasma cotinine concentrations after an adjustment for other factors than di& ad'olescents whose parents did not smoke. A plasma cotinine concentration >2.5 Fcglml was associated with an 8.9°lo greater ratio of total cholesterol to HDL cholesterol an& a 6.8% lower HDL cholesterol level (23). Similar results have been reported for nonsmoking adults exposed to ETS in the workplace (9). These results suggest that passive smoking, like active smoking, leads to alter*ations in lipid profiles predictive of an increased risk of atherosclerosis. The present study, however, showed no significant dif- ferences in total serum cholesterol, triglycerides, HDL- cholesterol, cholesterol-weeks, change in cholesterol, change in triglycerides or change in HDL cholesterol be- tween, the control group and the two passive smoking groups. To test whether the changes in~ lipid lesions associated with~ ETS exposure could: be a result of differences in cholesterol levels, we performed a multiple regression anal- ysis withe the percent aorta and pulmonary anery with lipid deposits 4 the dependent variables and cholesterol, triglyc- erides and HDL cholesterol7evels at 2 and 12 weeks (that is, before and after ETS exposure control) and ETS concentra- tion as the independent variables. In both cases ETS expo- sure was still significaa(p < 0:001) and positively associated with ETS dose after accounting for differences in serum cholesterol. Therefore, the increase in atherosclerotic le- sions in the cholesterol-fed rabbits exposed w ETS wass independent of changes in serum lipids in the present study. Passive smoking and platelet fnnction. In addition to their role in acute thrombus formation, platelets have also been~ implicated in the development of atherosclerosis. Davis et al. ('17) reported'that mean values of the platelet aggregate ratio before and after passive smoking were 0.87 and 0:78, respectively (p = 0.002): These valUes are similar to those we observed in the high ETS group (Table 3). They found that passive smoking increased platelet aggregation with a magnitude similar to that observed in active smoking. The effects of cigarette smoking on the levels of piatelet- activating factor (PAF), one of the most potent proinflaal- matory agents, or PAF-Gke lipids were studied (24,25). The .-N ~..~~

JACC Vol. '-1. No li January 1993.:25-3: results showed that, the values of PAF-like lipidsin both LDL and HDL in smokers increased sigrtificantly after smoking, and that the activiny of plasma enzyme, PAF acetylhydrolase. was inhibited by cigarette smoke extract in a dose-dependent manner. The charge alteration, reactions and'PAF-acetylhydrolase inhibition appear to be localized at differenb sites on the lipoprotein molecule. Thus, the ob- served inhibition may account for the increase in plasma PAF concentration that is known to occur in smokers. Inithe present study, bleeding times at week 12 in the two ETS groups, were significantly, shorter than in the control group (p < 0,001), and'platelet aggregate ratio at week 12 in the high ETS group was borderline lower than the control level (p = 0.07): suggesting increased platelet aggregate formation. These results suggest that the effects of passive smoking may be mediated. at least in part, by altered platelet function. Passive smoking and arterial endothelium: People ex- posed to ETS had a significantly thicker carotid artery wall than that of nonexposed persons who had never smoked, with the increase in wall thickness increasing with greater ETS exposure (8). Such epidemiologic studies are comple- mented by a variety of physiologic and biochemical data showing that ETS damages arterial endothelium. Davis et al. (17) reported that mean values of anuclear end'othelial cell carcasses in venous blood'before andafter passive smokingg were 2.8 and 3.7 (p = 0.002). The appearance of these cell carcasses indicates damage to the endothelium, which is the initiating step in the atherosclerotic process. Bondjers et al. (26) hypothesized that the effect~ of smoking might be medi- ated by increased catechol'amine levels. The endotieliall injury induced by smoking could be inhibited by metoprolol,, supporting this hypothesis. Other possible mechanisms of atherogenesis induced by F.TS. Clinical studies (27-29) in smokers with~ coronary aftery disease show that smoking increases myocardial oxygen demands and' such~ indicators as the rate-pressure product)- Also, smoking-induced': coronary vasoconstriction, which is due to an alpha-adrenergically mediated increase in coronary artery tone, is prevented by calcium antagonist drugs and nitroglycerin. Thus, smoking can adversely affect the balance between myocardial oxygen supply and demand. Several animal studies (5) have also shown that injections of polycyclic aromatic hydrocarbons, in particular 7,12- dimethylbenz(a,h) anthracene and benzo(a)pyrene, signifi- cantly increase aortic plaque and accelerate the develbpmentt of atherosclerosis. These studies provide evidence thatt known carcinogenic chemicals can be athcrogenic as wel1L In animal'4;xpPtiments, ETS also depresses cellular respiration at the level of mitochondria (30). The effects of ETS on c:urliov.tscular function, platelet function, neutrophili func- tii;n and plaque formation are the probable mechanisms le-t;r., ~,) he;ttt disease (4,5). Dose and duration. In the present study, the average concentrations of'air nicotine. CO and particles d'uring 7 h eF exposure in the high ETS group were 30-fold, 3-fold and zHU ET AL. '_3l' PASSIVE SMOKING INCREASES ATHEROSCLEROSIS 10-fold higher than in the low ETS group (1.040 vs. 30 µg!m3. 60 vs. 19 ppm. 13.8 vs. 1.2 mg/rrt3, respectively) Human exposure studies (5,11,13) showed thannicotine and respira- ble suspended particulate levels in restaurants ranged1from I to 25 Fcgim3 and 55 to 600 µgim3', respectively; respirable suspended particulate levels were 589 to 11,140 µg(m3 in bars an& bingo halls ('3), The U.S. National Ambient Air Quality. Standard for respirable particles is 50 µg.<m'' (annual aver- age). The nicotine levels in smoking sections on airplaneswere found to be 50 to 100 µg/m3. Air nicotine. CO and respirable suspended particulate levels in some public smoking rooms were foun& to range from 50'to 500 µg%m3, 5 to 50 ppm and 0.50-1.95 mg/m3; respectively. Thus, the levels we observedin the high ETS group are a factor of 2 to 10 higher than those observed in routine human environ- ments and the levels in the low ETS group are similar to those of heavily smoking-polluted, but realistic„ human environments. The studies (31) reviewed show that cotinine measure- ments are sensitive to the current exposure of nonsmokers to other people's tobacco smoke withia half-life of ='_0 h in the blood. Plasma cotinine levels after 2 hi of exposure to ETS in a heavily polluted public house were 7.33 ng.1m1. The cotinine levels in plasma we observed in the low ETS group were comparable to those of a heavily polluted room. whereas those in the high ETS gnoupiwere two- to fourfold' higher. Despite exposure to higher than routine human exposure levels, every rabbit in the two ETS groups tolerated the exposure well during the I0•week exposure penod. There were no differences in food consumption or weighn gain among the different' experimental groups. The differences between these experimental exposure levels and' actual humane exposure levelk were small compared with those of other.•studies of environmental toxins, where extrapolations >5 to 6 orders of magnitude are common. Indeed. our low ETS group represented realisvic exposure for people who work imsmoking environments, such as bartenders or wait- ers working in the smoking section of a restaurant. Conclusioas. These data indicate that the exposure of lipid-fed'rabbits to passive smoke adversely affects platelet function and significantly increases atherosclerotic lesions in the aorta and pulmonary artery. This increase imatheroscle- rosis is independent of changes in serum lipids and has a dose-response relation. These results are consistent with epidemiolbgic studies demonstrating that ETS increases the risk of death 1rom heart disease. Gratefuiappreciauon is given to James Repacc of the Envtroruacntal Protce- tion Agency, Waslungton: DC. wayne Ott of EPA and Sianford Ltuvcrsny.. tkpanment of Sutistics- Stanford, California and~ t.ee t.angan of, 1.aapn Producu. Inc. for, theu many thoughtful suggestions. We acknowledge John Hudson of the Electronic Facilities„Cardiovasculir Research Institute. l,'ru- versity of California; San Francisco. San Francisco, California for techuucal usistance. and Paul E. Sumner. San Franmsco for mva)uablc asslsunce In, the platekt aggregation studies. We alsoahanY S. Kathanne Hammond of the ~

t 4 232 ZHU ET AL. PASSIVE SMOKSNG INCRFASES ATHFROSCLEROS1S University of Massachusetts, Medical School; Worcester. Massachusetts for help in analying tbe au nicotine levels. References. 1., US Public Health Service. The health consequences of involUn- tary smoking: a report of~tlie Surgeon General. DHS1CDCl87-839'8, 1986., 2. National'Research Council. Environmental Tobacco Smoke: Measuring Exposure and Assessing Health Effects. Washington DC: National Aead: emy Press. 1986, 3, Repace JL. lndoor, concentrauons of environmental tobacco smoke: held survey. In: O'Neill 1K. Btunnemann KD: Dodet B. Hoffmann D, eds. Environmental Carcinogens. Methods of Analysis and Exposure Mea- suremenr, (Vol 9, Passive Smoking): Lyon: IARC Scientific Publications. 1987:141+59. 4. Taylor, AE. Johnson DC, Kizemi H. Environmental tobacco smoke and cardiovascular disease: a position paper from the Council on Canliopul- monary and Critical Care. American Heatt Association. Circulation 1992;86il-4. 5. Glantz SA. Parmley WW. Passive smoking and heart,disease::epidemiol- ogy, physiology„and biochemistry. Circulation 1991:83:1-12. 6. Steenland K. Passive smoking and the risk of heart disease. JAMA 1992 ~267:94-9 : 7. US Environmental Protection Agency. Respiratory health effects of passive smoking: lung cancer and other disorders. EPAl60fY69N006B.. May 1992. 8. Howard' G, Szklo M. Evans G- Tell G. Eckfeldt I., Heiss G. Passive smoking and carotid artery wall thickness: the ARIC study fabstrl: Circulation 199^_;85:8fi2. 9~ White IR, Cnqui M, Kulik JA. Froeb HF. Sinshelmer PJ. Serum lipoprotein in nonsmokers chronically exposed to tobacco smoke in the workplace. 8th World Conferonce on Tobacco or Health, Buenos Aires Argentina. March 30-April 3, 1992 (abstr): 10. Fisher ER, Wholey M.,Shoemaker R. Cigarette smoking and'cholesterol atherosclerosis of rabbits. Arch Padtol 1974:98:418-21. 11. Repace J, Lowrey AH. Indoor air pollution, tobacco smoke, and public health. Science 198+)'~208:464-72. 12. Ingebrethsen BJ, Heavner DL. Angel AL. Conner JM, Steiehen TI. Green CR. A comparative study of environmental tobacco smoke partic- ulate mass measurements in an environmenul chamber. J Air Pollut ControllA'ssoc 1988:38:413-7. 13. Hammond SK. Leaderer BP. A diffusion monitor to measurrexposure to passive smoking. Environ Sci Technol 198721 494-7: 14. Jacob P, Wilson M, Benowitz NL. Improved gas chromatographic method1or the determination of nicotine and cotiaiae in biologrc fluids. J!Chromatog 1981tZ'r2:61-70. 15. Jacob P, Yu L, Wilson M, Benowitz NL. Selected ion anottitoring method for detertninanon of nicotine, coeinine, and deutetium-labelled analogs: JACC Vol. 21L No. I Januarp 1993:)..25-32 absence of an isotope effect in the ciearrnce of (5)-nieotiise-3',3'd= in humans. Biol Mass Spectrom 19912-0:247-52. 16. Wu KK. Hoak JC. A new metDod for the quanutauve detection of plhtelet aggrngates in pauents with anenal insuffteiency. Lancet 1974:1i1:924-6: 17: Davis 1. Shelton L. Watanabe I, Arnold 1. Passive smoking affects endothellum and platelet. Arch lntern Med'1989:149:386-9, 18 . Wells A. An estiaute of adult mortality in the United States from passive smoking. Environ lat 19g8a4:249-65. 19. Kristensen T. Cardiovascular diseases and the work eovironmene:, a critical review of the epidertuologic literaturo on chemical factors. Scand J Work Environ Health 1989:15:245-64. 20. PDA.Y Research Group. Relationship of uherosclerosis in young men to serum bpoprotein cholesterol concentrations and smoking. JAMA 1990;, 264:3018-24. 21. Muscat JE, Harris RE, Haley NY. Wynd'er EL. Cigarette smoking and' plasnu cholesterol. Am Heart 1 1991':121:141-7. 22. Moskowitz WB, MosteUer M, Schieken RM, et al.. Lipoprotein and oxygen transport altenuons in passive smoking preadolescent childreo: the MCV'twin study. Circulation 1990:81:586-92. 23. Feldman 3, Shenker 1R. Etzel RA. et al. Passive smoking ahers lipid profiles in adoleseents. Pedtatncs 1991:88:259-64. 24, Ihuizumi T, Sa[oh K. Yoshida H, Kawamura Y. Hiramoto M. Takamatsu 5, Effect of cigarette smoking on the levels of platelet-acuvaung factor. like Gptd(s)' in plasma hpoproteins:. Atherosclerosis 1991:87;47-55:. ?S. Miyaura S! Eguchi H. Johnston JM. Effect of cigarette smokrextt9ct,on the metabolism of the proi.ttAammatory autatoid, platelet-activatutg fac- tort Cirr Res 1992;70;341-7. 26: Bondjera G, , Haasson G, Olsson G, Pettersson K. Smoking, cita cholamme and their effects on endotlielial cell integrity. lh:,Diana JN. ed. Tobacco Smoking and Atherosclerosis: Pathogrnesis and Cellular Mecb- anisms. New York: London: Plcaum Press. 1990:51-6: 27. Nicod P. RetinR, Winniford MD, CampbelliWB, Firth BG.,Hillis LD;. Acute systemic and coronary hetnodynantic and serologic responses to cigarette smoking in long-term smokers with atherosclerotic coronary artery disease. J Am Coll Cardiol'1 1984;4:964-71. 28. Winnifor,i MD, Wheelan KR, Kremers MS. et, al. Smoking-induced coronary vasoconstnction in patients with atherosclerotic coronary, artery disease: evidence for adtenergically mediated alterations in coronary, artery tone. Cirtulation 1986;7i:662-7: 29:, Winruford MD. Jansen DE. Reynolds GA. Apprill P, Black WH, Hillls LD:,Cigarette smoking-induced'eoronary vasoconstrieuon in atherosck- rouc coronary artery disease and prevention by calcium antagonists and nitrog.fycersn. Am I Cardiol 198759:203-7. 30. Gvozdjakova A, Bada V„Sany L, et al. Smoke cardiomyopathy: distur- bance of oxidative process in myocardial mitochondtia. Cardiovasc Res 1984;18:229-32. 31. Jarvis MJ.,App6ution of biocbemical intake markers to passive smoking measuremeot and risk estitnation: Mutat Res 1989'-l2.1:101-10.