U.S. DEPMtTMEWP OF HEALTH, EDUCATION, AND WELFARE. Smok+ and H&th: A Report of the Surgeon Geneml. DHEW Pub. No. 07960066. us. Department of Health, Education, and Welfare, Public Health Serviw, offi~ e the h-t ,%mtmy for Health, Offhx on Smoking and Health, 1979. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVLCES. The Health am ~a- of &noking: Cancer. A hfq.wt of the &q&on &neml. DHHS pub. &. (pHs)8250179. U.S. Department of Health and Human Services, Public I&&b &IV&, office of the Assistant Secretary for Health Office on Smoking w Health, 1932. U.S. DEPARTMENT OF HEALTH AND = s-vK.!E% !t"he Health &- ~uence.g of Smokingz Chmnic Obstrueti~ Lung Disease A Repor of the Sugcon Ged. DEIHS Pub. No. @%S)34-50205. U.S. Department of Health and Hm &WV&S, public Healtb Service, office of the Assistant Secretary for Health, OfIiee on Smoking and Health, 1934. U.S. DEPARTMENT OF TRANSPORTATION and U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Health Aspect.9 of Smoking in 32ane port Aitwaf% U.S. Department of Transportation, Federal Aviation Admi&tm- tion, and U.S. Department of Health, Education, and Welfare, National Inatit& for Occupational Safety and Health, December 1971. U.S. FED- TRADE COIvfIKMION. "Tar, "Nicotine and Gzrbon lldonozide of tke Smoke of 207 V&i&es of Dm&ic Cigarettes. U.S. Federal Trade Comm%m, January 1935. VAINIO, H., HEMMINRI R, WILBOURN, J. Data on the carcinogenici~ of chemicale in the IARC ddonogmphe p rogramme. Cunzinogenesis 6(11~1653-1665, November 1935. VILCINS, G., LEPHAIUYI?, J.O. Aging procewea of cigarette smoke: Formation of methyl nitrite. C&m&y and I&u&y 223974-975, November 15.1975. WADE, W.A. III, COTE, W.A., YOCUM, J.E. A study of indoor air quality. Joumai of Ule Air Pollution Control Association 25(9):93%939, September 1975. WALD, N., RITCHIE, C. Validation of studies on lung cancer in nonmokers married by .sders. (letter). Lamt 1@335):607, May 12,191~. WALLAcE,UpELLIzzARL E.D., HARWEL, T. SPRACINO, C., ZEI.QN, H. Personal exposures indoor-outdoor relationships and breath levels of volatile Ogaaics in New J-y. Atmospheric Envinmment, in press. WALSII, I% BLAcg, A., MORGAN, A., CF&%WSHAW, OH. Absorption of SO, by tppical indoor sllrfacee including wool carpets, wallpaper, and paint. Aanaepheric J%virowieent 11(11):1107-1111,1977* 176 CHAPTER 4 DEPOSITION AND ABSORPTION OF TOBACCO SMOKE CONSTITUENTS CONTENTS Deposition Size Distribution of Cigarette Smoke Mainstream Smoke Sidestream Smoke Particle Deposition in the Respiratory Tract Total Deposition Regional Deposition Respiratory Tract Dose of Environmental Tobacco Smoke Cigarette Smoke Particulate Mass Deposited The Concept of "Cigarette Equivalents" Markers of Absorption Carbon Monoxide Thiocyanate Nicotine C&nine Urinary Mutagenicity Populations in Which Exposure Has Been Demonstrated Experimental Studies Nonexperimental Exposures Quantification of Absorption Evidence of Absorption in Different Populations Quantification of Exposure Comparison of Absorption From Environmental Tobacco Smoke and From Active Smoking Conclusions References Introduction An understanding of the deposition of &arette smoke partick in the respiratory tract is important because many of the toxic cmstituents of cigarette smoke are conbined in the particles. me quantity retained, which c~nstitutea the dose, is some fraction of the quantity inhaled. Meas~reg of tobacco smoke constituents or their metabolites are also important because they reflect the absorption of tobacco smoke by the individual smoker or nonsmoker, and therefore may be more accurate markers of the actual exposure experienced by an individual. There is little experimental information describing the deposition of environmental tobacco smoke in the respiratory tract (Jarvis et al. 1983). However, cigarette smoke particles probably behave in a manner similar to other inhaled particles. I,n contrast, there are a number of observations of different markers in the biological fluids of smokers and nonsmokers. This review begins with a discussion of particle deposition in general and the factors that affect deposition. This understanding is then applied to the existing data on tobacco smoke deposition in the human respiratory tract. Subsequently, a variety of biologic markers of smoke absorp- tion are e namined, and the levels of these markers found in smokers and nonsmokers under a variety of circumstances are presented. ~inal.ly, an attempt is made to qua&it& the exposure of nonsmok- ers relative to that of active smokers using levels of these biologic markers. Deposition The term "deposition" refers to the transfer of a particle from inhaled air to the surface of any portion of the respiratory tract, from nose to alveolus. "Retention" is the quantity of deposited material remaining in the respiratory tract at a specified time following deposition. Retention decreases as clearance mechanisms such as mucociliary action and absorption reduce the respiratory tract burden of inhaled particles. Retention is not discussed in this review. An aerosol is a suspension of particles in a gaseous or vapor medium; cigarette smoke is an aerosol. Aerosols are characterized by such terms as mass median diameter @MD), the diameter below which lies one-half of the particles by mass, and count median diameter (CMD), the diameter below which lies one-half of the particles by number. Most naturally occurring aerosols have a log- normal size distribution, and the magnitude of the spread of particle size is the geometric standard deviation @SD). Particle ma88 is a function of the cube of the diameter; a particle with a diameter of 0.5 pm has one one-thousandth of the mass of a 5 pm particle. Thus, for au aerosol with a large geometric standard deviation, the mass 181 median heter may be considerably greater than the count median hemr. The smaller pdicle~ of an aerosol, despite tbeu relatively small mass, have a large total surface area because of their great number. A monodisperse aerosol has particles of one size, with C&ID equal to M&ID, and a G-SD of 1. For practical purposes, a GSD of 13 or less is accepted as monodisperse. Most naturally oawring mrmb EU-C? p~lydisperse, with GSDs in the 2 range. A lognormally &t&&cl aerosol with a GSD of 2 and a CMD of 0.1 will have an MMD of O.&-Z. In this discussion, when size is referred to, it is the JMI$D unless otherwise stated. Both the total deposition and the `deposition site in the respiratory tract vary substantially with particle size. site wtributlon of Cigarette Smoke m Smoke The size distribution of cigarette smoke hes been of interest to investigators for many years. The important relationship between size and respiratory tract deposition is discussed below. Most studier have been performed using mainstream smoke. Mainstream smoke is the smoke exiting from the butt of the cigarette during puff- drawing, and sidestream smoke is the smoke plume that drifts into the environment from the burning tip of a cigarette between puffk. Environmental tobacco smoke @TS) is the ambient burden of side&ream smoke and the smoke exhaled by a smoker. Involuntary smoking is the consumption of ETS by people, either smokers or nonsmokers, from the environment. One purpose in discus&g the size distribution and respiratory tract deposition of particles is to illustrate the discrepancy between the measured particle size of mamstresm smoke and its measured deposition in the human respiratory tract. The deposition fraction of mainstream smoke is several times higher than would be predicted on the basis of its particulate size. The measured deposition of side&ream smoke is more in keeping with its measured size (Hiller, McCusker et al. 1982). The standard laboratory smoke-generation technique is tc force air through the cigarette as would be done by a smoker, followed by the rapid dilution of the resulting mainstream smoke so that particle size can be measured. A standard 36 cma, 2-second puff is usdY used, although actual puff volume was shown to average 45 cm' in one tidy (Mitchell 1962) and 56 cm3 in mother; for individuals, the Puff vohme can vary from 20 to 30 ems UP to 70 to 80 ems (Hinds et al. 1983). The sire distribution of the diluted mainstream smoke aerosol in then m-u& by one of a variety of techniques such as light scattering devices, microscopic measurement, or impactor collectipg 182 devices. using various diluting and sizing techniques, particle s& measurements of mainstream cigarette smoke have been repor-teo from many laboratories (Table 1). One potential cause of error in measuring the Sk%3 distribution of main&-e&m cigarette smoke & the relative insensitivity to ultrafine particles of some previously used measurement methods. More recent studies using newer measure- ment techniques support the suggestions by the earlier investigators (Sinclair 1950) that there is an ultrafine, (< 0.1 pm) component u-~ the cigarette smoke. Size characteristics have been measured by electron microscopic methods, following rapid fixation of undiluted fresh tobacco smoke, as CMD 0.2 pm and GSD 1.5 (Keith 1982). me sb distribution measured with an electrical aerosol analyzer has been reported as CMD 0.1 pm, GSD 2.0, suggesting more ultrafine particles than previously recognized (Anderson and HjJ,ler 19%). Smaller particles (< 0.4 pm) of tobacco smoke have been shown to have a chemical composition different from that of larger particles (St&x 19&I), possibly because of the large surface area of smaller particles. Laboratory methods, such as rapid dilution, commonly used to study mainstream smoke, are highly artificial and may not accurate- ly duplicate the generation, dilution, and inhalation of mainstream smoke by the smoker. Smoking technique and respiratory tract conditions may promote changes in particle size. Therefore, the particulate sizes in the respiratory tract may differ from the sizes measured when mainstream smoke is diluted for size analysis or when diluted sidestream smoke is inhaled by the involuntary smoker. The smoker's puff is taken as a bolus in a relatively small volume of air into the humid upper respiratory tract. Smoking techniques vary widely (Griffrtbs and Henningfield 1982) and have been shown to vary significantly among groups classified as healthy smokers compared with those with emphysema and also between those with emphysema and those with bronchogenic carcinoma and bronchitis (Medici et al. 1985). Some smokers hold the puff in the mouth for several seconds prior to deep inhalation. The initial puff is highly concentrated, with approximately lo8 particles/ems. At this concentration, particle coagulation can occur rapidly, causing a tenfold to a hundredfold reduction in particle number and an increase in particle size (Hinds 1982). Also, the accumulation of water in or on the particles in the high humidity of the respiratory tract can increase particle diameter (Muir 1974), and may increase the diameter as much as 30 percent (Mitchell 1962). Some evidence suggests, however, that at least for dilute cigarette smoke, hygrc+ scopic growth occurs only under supersaturated conditions (Kousaka et al. 1982). Coagulation and water uptake by particles in the respiratory tract may considerably alter particle size distributions so that measurements under laboratory conditions probably do not 183 !i `l'ABLE L--Size distribution of mainstream tobacco smoke Study Size (pm). concentration [no. particles/cm'] Dilution Method Comment Wells and Gerke (1919) Sinclair (1950) DallaValle et al. (1959) CMD 0.27 Not given Oscillation amplitude CMD 0.0-0.3 fresh CMD 0.4-0.5 aged 0.1-025 Not given Light scattering Electrostatic separation Agedz size increase attributed to water accumulation Langer and Fisher m53 CMD 0.5 filter CMD 0.6 plain 12-6 x 1OY 1431 Microscopic impinger collection Compared with electrostatic precipitation mm 1.75 Keith and Derrick (1960 CMD 0.23 MMD 0.46 29&l Aerosol centrifuge Mic-pio GSD 1.64 C.dOUl&!d Porstendijrfer and ChlD 0.22 Schraub (1972) [5-7 x 108] 1OO,ocQ1 Related rate of deposition of radioactive decay products onto particles to particle 8ir.e Also measured deposition Pomtendiirfer (1973) CMD 0.42 CMD 0.22 lo:1 3,loo:l Radon daughter attached and deposited in spiral centrifuge Okada and Matsunuma (19741 CMD 0.16 MMD 0.29 1,50&l Liiht scattering GSD 1.46 TABLE l.-Continued Size (pm), concentration Study [no. particleelcm'] Dilution Method Comment Hinds (1978) McCusker et al. (19821 Chang et al. MMD 0.38-0.62 CMD 0.4 CMD 0.27 MMD 0.29-4.3 [4.2 x lO*] lO:l-7oo:l lo:1 3,loozl 126,000:1 Aermol centrifuge Laser doppler v&&o&y Size distribution decreases as dilution incrensee GSD 1.3-1.6 Aerodynamic diameter OSD 1.4 CMD 0.24-0.26 [3.6 x lay MMD 5.5 secondary mode &l-18:1 Electrical aerosol aaaIyzer (EAA) Anderson Cascade Impactor CD Bimodal distribution ww l-8 I 100 Primary mode @AA) CSD 1.18 Secoad mode (CD Sk-3046 of total mass NOTE: CMD = count median diameter: MMD = mass medii diimeter; GSD = geometric standard deviation. TABLE 2.-&e distribution of sidestream tobacco smoke StudJr size (pd DiIUtkUl comnl~t Keith and Derrick w6u CMD 0.16 296zl AWSOl Nature of sideetxeam centrifuee centrifuge smoke generation pmcmamakeadif6cult exact detednation of wncentration at generation and dilution Pol-sten&rfer and Schraub (1972) CMD0.24 Not given Rewed rate of depceition of mdiwctive decay products onto prticIee taparticIe&e Hiller, Mecusker et al. w32) CMD 0.31 Not &en Laser doppler veIwiIn* GED 1.6 NUlECMD=axmtmediMdiamaa;~=geometric etamid deviation represent distributions found in actual mainstream smoking condi- tions. Side&ream Smoke Sidestream smoke is generated by cigarettes burning spontaneous- ly between puffs and is quantitatively the major contributor to ETS. Fit+five percent of the tobacco in a cigarette is burned between puffs, forming sidestream smoke (see Chapter 3). Dilution takes place as smoke rises in the ambient air currents. This dilution with air reduces, but probably does not eliminate entirely, the coagulation that causes the particulate to increase in size, as they may in the highly concentrated state that occurs when a smoker draws a puff of mainstream smoke into the mouth and holds it briefly before inhalation. The size distribution of sidestream smoke might be expected to resemble that of diluted mainstream smoke. The results of several reports of sidestream smoke size measurements (Table 2) support this impression. Particle Deposition in the Respiratory Tract Total Deposition Total deposition haa been studied both theoretically and experi- mentally. Mathematical equations can be used to predict deposition by combining mathematical models of lung anatomy with equations describing the behavior of particles in tubes. The major property to be considered is particle size and its influence on impaction, sedimentation, and diffusion. Inertial impaction is the mechanism 186 that causes Particles moving in an airstream to be tile, because of exceesive mass, to follow the airstream around a bend. Large particles impact at the bend in the aimtmam or m the lung on or near a site of airway branching. The Iarger the particle the greater its chance of depositing by impaction. Impaction ia a relatively unimportant form of deposition for particles smaller than 0.6 pm. The effect of gravily on suspended particles causes them to fall, a process called sedimentation, which also becomes relatively unim- portant for particles less than 0.6 I.Lm in size. Larger particles fall faster, end for all particles, the greater the residence time (in the lung) the greater the likelihood of deposition by sedimentation. Diffusion is the net transport of particles caused by Brownian motion. It becomes increasingly important for particles less than 0.5 pm in size (Hinds 1982). The mass median diameter of aide&ream smoke is in the 0.3 to 0.5 w size range. Total deposition for inhaled particles is in the 10 to 30 percent range for 0.5 pm sixed particles. In Figure 1, Lippmann's review (1977) of the measurements of total deposition of monodisperse aerosols in human subjecta is modified to include more recent data and data on ultrafine particle depcsition. The respiratory pattern clearly affects particle deposition. Most important for all particles, including environmental tobacco smoke, is the residence time in the lung. Deposition increases with slow deep inspiration (Altshuler et al. 1957) and with breath holding (Pahnes et al. 1966, Anderson and Hiller 1985). In hamsters, the deposition of 0.38 pm particles rises in a nearly linear fashion with oxygen consumption (Harbison and Brain 1983). These data indicate that deposition of ETS during involuntary smoking increases with the increasing activity level of the exposed individual. The presence of an electrical charge on particles may increase deposition, Mainstream smoke is highly charged (Corn 1974). The addition of either a positive charge or a negative charge to inhaled particles increases deposition in animals (Fraser 19&l), and neutral- ization of the charge reduces deposition 21 percent in rats (Per-in et al. 1983). There is little information describing the effect of a charge on the deposition of either mainstream or side&ream smoke in human subjects. Particle growth by water absorption may affect deposition. Mathe- matical models that describe the effect of humidity on particle growth indicate the potential for a considerable change in size of some particles during transit in the humid rt -iratory tract (Perron 1977; Cocks and Fernando 1982; Renninger et al. 1981; Martonen and Pate1 1981) and that these changes could ~igniikmtly alter deposition @`err-on 1977). Growth of 0.4 to 0.5 p particles should increase their deposition fraction, but growth of a 0.07 pm particle to 0.1 pm, for example, would reduce its deposition (see Figure 1). Such 187 I I I 0.01 0.02 0.03 0.05 0.1 0.2 0.3 0.5 0.7 1.0 2 3 5 7 10 20 Aerodynembic dim&r-pm an effect has been shown for laboratory-generated aerosols in human subjects (Blanchard and Willeke 1983; Tu and Knudson 1984). While hygroscopic growth has been postulated for tobacco smoke (Muir 1974), it has been demonstrated in the laboratory to occur, at least for dilute smoke, only in supersaturated conditions (Kousaka et al. 1982). Many reports describe measured deposition of mainstream ciga- rette smoke in the human respiratory tract CTable 3). Although few studies of total sidestream smoke deposition are available, those few (Table 3) suggest that sidestream smoke does indeed deposit in a manner similar to that found for laboratoryde&ned research aerosols. The deposition fraction of mainstream smoke diluted 1:30 and inhaled by rata from chamber air containing 1.68 mg/L (assuming a rat tidal vo 1 ume of 1.5 mL and a respiratory rate of 85) is 138 8.1 percent @IUU et d. 1978). &positiop for the side&ream smoke has been measured in mouth-breathing human volunteers at 11 pertint, similar to that for similarly sized polystyrene latex spheres (Hiller, Mazumder et al. 1982). Environmental tobacco smoke exposure frequently occurs with breathing through the nose rather than through the mouth, but inert particles in the size range of ETS (0.2 to 0.4 pm) are not substantially reduced in number by passage through the nose. The fraction of inert 0.2 pm particles deposited in the alveolar region of the lung is similar for mouth breathing and nasal breathing (Raabe 1984). It is possible that the charged or reactive particles of J3TS may behave somewhat differently than inert particles, but it seems unlikely that nasal breathing substan- tially alters the deposition of the small particles of EY.CS in comparison with mouth breathing. Ftegiomd Deposition Total deposition is subdivided into the fractions depositing in the upper respiratory tract (larynx and above), the tracheobronchial region (trachea to and including terminal bronchioles), and the pulmonary region (respiratory bronchioles and beyond) (Figure 2). Deposition in these areas is referred to as regional deposition. Particle size is a major dete rminant of both total and regional deposition. A mathematical model prediction of regional deposition of polydisperse aerosols is shown in Figure 2 (ICRP 1966). Experimental verification of mathematical models of regional deposition is limited. Using isotope-labeled particles, it is possible to quantitate the upper respiratory tract deposition as a fraction of total deposition. By assuming that the aerosol depositing in the tracheobronchial region will be cleared within 24 hours, it is possible to measure alveolar deposition as the fraction of the total initial deposition below the larynx that is remaining at 24 hours and tracheobronchial deposition as the difference between the initial deposition and what is remaining at 24 hours. Using this method, the deposition of 3.5 pm particles was this: total deposition, 0.79; upper respiratory tract, 0.10; tracheobronchial region, 0.24; and pulmonary region (alveolar), 0.45 (Emmett et al. 1982). These measurements are below the estimated regional deposition for upper respiratory tract deposition and higher for the pulmonary deposition than are the measurements calculated by using the Task Group on Lung Dynam- ics model (ICRP 1966). `The regional deposition of mainstream cigarette smoke in smokers ha alSo been studied. Subjects inhaled smoke from cigarettes labeled with radioactive l-iodohexadecane (Black and Pritchard 1984; Pritchard and Black 1984). The results indicate that less than 40 percent of the particulate mass deposited in the pulmonary region, compared with an expected 90 percent deposition in the @I TABLE 3.-R aspiratory tract deposition of mainstream and aide&ream cinarette IWWSQ Study Mainstream smoke Deposition fraction Puff volume CmL) Smoke dilution Respiratory pattern Beumberger (1923) Schmahl et al. (1954 Polydorova (1961) Mitchell (1962) Dalhamn et al. U963) 88% 98% 80% (22-39 range) 32% (70-90 range) 98% + 3.1% SD @6-99 range) Not given 46 f 9.8 SD (33-65 range) 3s Puff time &and) Not given 1.9 f 0.6 SD 2 None None 3OOZl None Hinds et al. 41% m33) (22-75 range) 53 None Inhalation Usual spontaneous smoking pattern "Deep inhalation" Pretrahled shdard~ pattern (not dencried) Ueual epontaneoua smoking pattern Side&ream smoke Binns et al. (1978) Hiller, McCusker et al. (1932) 8% 11% Not applicable Not applicable 3o:l (in cbarnber~ 5I-100 palm' Spontaneous (rata) 1 L tidal volume, 12 breatheJmin Figure Z.-Regional deposition of particles inhaled during nasal breathing, as predicted using the deposition model proposed by the Tack Group on Lung Dynamics pulmonary region for 0.5 pm particles, the size reported for cigarette smoke (Table 1). This finding further supports the concept that mainstream smoke particles increase in size in the respiratory tract by coagulation, hygroscopic growth, or both, and that this growth affects total and regional deposition. The same group studied the effect of switching the tar content of cigarettes on regional deposi- tion. Using. cigarettes with between 16 and 17 mg tar, extrathoracic deposition was found to be 14 percent of the total deposition and intrathoracic deposition to be 86 percent, with 51 percent in the tracheobronchial area and 35 percent in the pulmonary region (Pritchard and Black 1984). After switching to cigarettes with between 8 and 9 mg tar, total deposition was 74 percent of that measured from cigarettes with the higher tar content, the extratho racic deposition was unchanged, the tracheobronchial deposition was from 34 to 42 percent, and the pulmonary deposition was 18 to 25 percent of the total mass deposited with the higher tar cigarettes. With the use of mathematical deposition modeling, the observed deposition pattern was consistent with one predicted for an aerosol with an MMD of 6.5 pm, more than 10 times greater than the MMD described for cigarette smoke (Black and Pritchard 1984). The deposition of particles is probably not uniform within a lung region. The mass deposited in the airways, for instance, may vary 191 widely. -4 deposition at specific anatomic sites may ba eswy important for mme inhalants. For example, the concentra- tion of carcinogenic subetances at a site may favor that site for cancer development. This may be e@ecmhY hW"rtant for c*arett@ smoke, hm lung mcer my occur at sites of high deposition such 88 my bifurcations. &p&ion of a 0.3 v laboratory-generated stable -1 has been shown to favor right upper lobe deposition, and on t-he ba& of surface density of deposition, the lobar bronchi (wesmger and Lippmann 1978). The deposition per airway genera- &n haa been d&~ for large particles, but has not received sufficient a-eon for p&idea in the size range of main&ream or Bidegtream smoke. A deposition peak has been Predicted, using a lung model for the fourth airway generation (trachea is 0) for 5 q particles, and a peak in airway surface concentration density was predicted for 8 w particles at the fourth generation (Gerrity et al. 1979). Both of these deposition peaks are calculated for particles &&&ially larger than those of cigarette smoke. )3q0&ions may be quite nonuniform even within a single airway generation. An enhanced deposition at bifurcations with highly concentrated deposition on carina ridges within bifurcations has been demonstrated in a five airway generation model of the human respiratory tract for both cigarette smoke (l&u-tonen and Lowe 1983a) and restxnch aerosols (Martonen and Lowe 1983153. Epidemiological studies of the pathophysiologic consequences of involuntary smoking have emphasized, among other things, an increase in the incidence of respiratory illness in children (see Chapter 2). The issue of the respiratory tract deposition of particles in children has been addressed only recently. Using morphometrm measurements from casts of the lungs of children and young adults aged 11 days to 21 years, a mathematical growth model was created. Using this model and conventional methods for predicting the behavior of particles in tubes, the deposition of particles at various ag- c-m be presided. On the basis of these calculations, tracheo- bronchial depositions per kilogram of body weight for 5 p particles Was &h&d b be six times higher in the resting newborn than m a resting adult @halen et al. 19851. Differences are predicted &o for particles the s&e of sidestream smoke, with tracheobroncm deposition m infancy beii twofold to threefold higher in adulthood. Total deposition has also been estimated using mathematical model- ing, with the total deposition estimated at approximately 15 percent at age 6 months and at 10 percent in adults (XU and YU 1936). 192 ResphtoW Tract Dose of EnvIronmental ~-cc0 smoke Cigarette Smoke Particulate lbht3 ~epoeited The dose of environmental tobacco smoke to the respiratory tract is the product of the mass in inhaled air and the deposition fraction. To this point, particle size and deposition fraction, which is related to both size and respiratory pattern as well as to other less understood factors such as particle charge and hygroscopicity have been addressed. To estimate dose, the content of smoke in inhaled air- must be known, as well as the respired minute volume. m content in inhaled air varies widely, as does minute volume, which depends considerably on activity level. Sidestream smoke concentrations have been raised as high as 16.5 mg/m5 in expetiental chambers (Hoegg 1972). High levels, 2 to 4 mg/ms, have also been estimated using measured carbon monoxide concentrations for rooms 146 m' in size containing 50 to 70 persons @ridge and Corn 1972). Such levels far exceed the EPA air quality standards for total suspended particulate of 75 w/m* annual average and the 260 w/m* 24-hour average in the United States and the 250 w/ma 24-hour average for the United Kingdom. Measurement8 of environmental smoke concentrations vary wide- ly, depending upon the location and measurement technique (Tables 4 and 5). Levels of total suspended particulates (TSP) measured under realistic circumstances have been found to be from 20 to 60 w/m3 in no-smoking areas, and can range from 100 to 700 &ms in the presence of smokers (Repace and Lowrey 1980). These measure- ments include all suspended particulates, and so could include part&e other than tobacco smoke. However, in a smoky indoor setting where measurements as high as 600 w/m9 have been found, tobacco smoke is the major contributor to particulate mass, with the non-tobaccwmoke contribution being small and similar to that measured for nonsmoking areas, namely in the 20 to 60 @mS range. This concept is supported by studies in which tobacco smoke concentration in the environment was determined by measuring the Gcotme content of suspended particulates. Using this technique (Hinds and First 1975), EI'S levels have been ehimati to be 20 to 4.80 p,g/ms in bus and airline waiting rooms and as bigh as 640 p&m" in cocktail lounges. These calculations of smoke concentrations were based on an average weighted nicotine fraction of 2.6 percent, an approach that may underestimate tobacco smoke particulate concen- tration. The mass deposition in the respiratory tract can be estimated if the atmospheric burden of cigarette smoke particulates, minute volume, and deposition fraction is known. Assuming a smoke concentration of 500 CLg/mS, a minute volume of 12 liters per minute, 193 TABLE 4.--Indoor concentration of total suspended p&cUlates (TSP) meamlred iu OIY%II~ living or work.ing situations Study Just et al. (1972) Hinds and First (1975) Location Coffee shop Conditions of location, owupancy, rmloking (5). nonsmoking @IS) 4 locations TSP pm/m' x &SD 1,160 Rackground pm/m* 670 ' Comments Bus waiting room Restaurant Cocktail lounge Arena A Arena B Arena C Not given Not given 40 m-68 200 (51-450) WE40) Not applicable Suspended paarticulates _ _ collected on filter; nicotine content measured for calculation; TSP = nicotine/O.026 Elliott and Rowe (1975) Attendance 9,660 Air conditioned (S) Attendance 14,300 Air conditioned (S) Attendance 2,900 Not air conditioned 9X Attendance 11,990 Natural ventilation (NS) 224 42 461 42 620 92 143 71 High volume sampler for suspended particulates; also measured Co at all locations and benr4alpyrene in arena A Cuddeback et al. (1976) Tavern 6 air changes/hr Tavern None apparent 0.31 f 0.05 (0.23-0.34) 0.99 8hr air sample collected on filter (6 pm pore size); TSP measured gravimetrically Neal et al. (1976) Hospital intensive care units Independent ventilation 30 66 Anderson personnel sampler systems ussd TABLE 4.-Continued Study Location Conditions of location, occupsncy, smoking 03, nonsmoking (NS) TSP pm/m" x &SD Sackground pm/m3 Comments Weber and Fischer (1930) Repace and Lowrey (1980) 44 offices Residences Libraries, churches, restaurants Restaurants, bars, bingo game Window ventilation; 32144 allowed unrestricted smoking Air conditioned 5 locations, 6 measurements, 10 * 8 persons/l00 m3, all NS 9 locations; 10 f 10 perwns/100 m', all NS 19 locations, 20 samples, 11 + 8 persons/100 ms, all S locations 7 locations with >l smoker/ms (mean 2.2 smokers/mJ) 18 f 7 persons/lo0 ma, with 1 smoker/l00 ma ,202 120 38 f 16 38 f 16 242 f 176 m-697) 406 * 188 (187-697) Subtracted from TSP Same Not done 36 + 10' (4 locations) 47 f 13' (13 locations) 53 f 8' TSP measured with pieeoefectric balance (see above) All samples colfected using pieroelectric balance with very high collection efficiency at 3.6 pm and 10% at 4 (cm; sample the l-60 min, outdoom 6-16 min z TABLE 4.-Continued Study Location Conditions of location, occupancy, smoking 6% nonsmoking (NS) TSP pmhn' I fsD Background pm/m' Comments Spengler et al. (1981) 35 homes 16 homes 5 homes 1 home* No smokers 1 smoker 2 smokers 2 smokers, tightly sealed, central air conditioning 24.4 f 11.6' 36.6 f. 14.6 70.4 f 42.9 144 21.1 zlz 11.9 all 55 homes Annual mean: respirable ma8s collected on filters after removal of nonrespirable fraction; `24-hr sample collected every 6 days ' Ambient prticulate concentration at site, but outdm~~. `This home is one of the !ive homes above. TABLE &-Indoor concentration of total suspended particulates (TPM) generated by smoking cigarettes under laboratory conditions chamber Cigarette TPM Study Test conditions Ventiition Size consumption mg/m' Comment.9 Penkala and Well mixed None 9.2 ma 3 simultan~usly, 2 q 3.0 de Oliveira (19'76) puffs Hoegg Sealed chamber; Portable fans 25 ma 24 simultaneously by 16.66 (1972) experimenter and test circulated air TPM measured gravimetrically machine after collection of suspended equipment in chamber; particulatea on fdten; measured 18 min sidestream smoke mlIected in pcetamoking chamber; mainstream smoke dischaqp?d Same, 150 min Same 4 simultaneously by 1.61 wetsmokina machine Hugod et al. (1978) Sealed room Unventilated 68 mJ 20 simultaneously by machine 6.76 TPM measured gravimetrically from 3hr collection on filter; mainstream smoke in chamber Cain et al. (1983) Muramahm 4-12 occupants Climata-controlled chamber Climatec0ntr0lled 11 ft3/min/oceupant 11 ma 4/hr (by occupants) 0.350 66 ft'lminloccupant 11 mJ 4/hr fby occupants) 0.16 11 ft'/min/occupant 11 ma 16/hr (by occupants) 1.26 68 ft'/min/occupant 11 mJ 16/hr (by GCGU~MW 0.40 16.4 air changes/hr 34 ms l/8 min to 60 min 0.19-0.26 F5ezoelectric balance messwed total mass over 0.01-20 pm Pieaoelectric balance et al. (1963) chamber Climate-controlled chamber 16.4 air changeslhr 30 ma 3 simultaneously, then 2/8 min 0.47-0.622 and a deposition fraction of 11 percent (Hiller, McCusker et al. 1982), mass deposition over an &hour work shift would be 0.317 mg. The Concept of "Cigarette Equivalents" Many investigators have attempted to estimate the potential toxicity of involuntary smoking for the nonsmoker by calculating "cigarette equivalents" (C.E.). To inhale one C.E. by involuntary smoking, the involuntary smoker would inhale the same mass quantity of ETS as is inhaled from one cigarette by a mainstream smoker. This approach has led to estimates from as low as 0.001 C.E. per hour to as high as 27 C.E. per day (Hoegg 1972; Hinds and First 1975; Hugod et al. 1978; Repace and Lowrey 1980). These differences of up to three orders of magnitude seem illogical when most reports of measurements of environmental concentrations of smoke, from the most clean to the most polluted with environmental tobacco smoke, are within tenfold to fiftyfold of each other. The following discussion demonstrates why the C.E. can vary so greatly as a measure of exposure. The calculation of C.E. is as follows: PMIw = TSP (mg/m') x Ox; where PM&,) equals the particulate mass inhaled by passive smoking, TSP equals the total suspended particulate, and VE equals the inhaled volume. C.E. = PMI&PML); where C.E. equals cigarette equivalent and PML) equals the mass inhaled by (mainstream) smoking one cigarette. (This is taken to be the tar content of a cigarette as reported by the U.S. Federal Trade Commission.) Cigarette equivalents can be calculated for any time interval chosen, i.e., per hour, per day. Although the example given is for particulate mass, C.E. can be calculated for any component of cigarette smoke, such as carbon monoxide and benzo[a]pyrene. The following calculations illustrate the different results from two different approaches to the calculation of C.E. Example 1 Example 2 SIB 0.36 mg/hr 20 ma/day PMIw 16.1 mg tar/@ 0.55 mg tar/cig TSP 40 CLg/mS 700 pg/m9 Example 1 PMIcp, = TSP x Ox = 40 P /m9 x 0.36 ms/hr = 14. CLg/hr C.E. = PMI&PM&ms, = (0.0144 mg/hr)/(l6.1 mg/cig) = 0.001 cig/hr 198 Example 2 PM&PI =TSPxOr = 700 w/ma x 20 ma/&y = 14,990 p&day C.E. = PMldwIk4 = (14 mgklay)/(O.65 mghig) = 245 &g/day These caktitiOn8 of C-E. approximate the approach- d h ho ~~rts-Exanw~e 1 by Kinds and Pi.& (1975) and Example 2 by Repace and Lowrey (198O)-and the results are similar. The exam- ples are the extremes used in the two studies, and are at the extremes Of CommOmy cited rep&a of C.E. Even if the Tsp concentration used in the two examples were the me, the multx would differ 24-fold because Example 1 is calculated per hour and Example 2 is calculated per day; 2%fold because of the aerence in inhaled minute volume; and 29-fold because of the difference in what is considered to be a `Wandard" cigarette. Even using the same TSP concentration, the results would be 1.6 x 10" different. If C.E. is to be calculated, all of the factors used in the calculation should be Standardized. The calculation of C.E. is deficient in several other ways. The deposition fraction of the total inhaled particulate mass in the respiratory tract from mainstream smoke is higher than from involuntary smoking. The deposition fraction for involuntary smok- ing is approximately 11 percent for mouth breathing (Hiller, Mazumder et al. 1982). The deposition from mainstream smoke has been reported to vary from 47 to 90 percent (Table 3). The cigarette equivalent calculation considers only the quantity inhaled, and if mm dose depoeited is considered, one C.E. from passive smoking will cause several times less mass to be deposited than the mainstream smoke of one cigarette. The differences in the chemical composition between sidestream smoke and mainstream smoke make the C.E. concept misleading unless C.E. is calculated for each smoke constituent. This has been accomplished (Hugod et al. 1973) using measured levels of various smoke co&ituente in a chamber filled with sidestream smoke. The redts indicate that one C.E. for carbon monoxide could b i&&j 5.5 times faster, and for aldehyde, 2.9 times faster, than for particulate mass. Measurements of total particulate matter and benxc(a]pyrene taken in an arena with active smoking revealed a fivefold rise in TSP above background and an eighteenfold increase in benzo[ajpyrene over background. Using the measured ben- zo[alpyrene concentration of 21.7 ng/ms, an inhaled volume of 2.4 ma, and 3.2 ng benxo[ajpyrene per cigarette, the occupant of such an environment would consume 6.4 C.E. for benzo[ajpyrene (IARC 1986, p. 87). The C.E. TSP would be 1.7. Therefore, a C.E. for the 199 carcinogen bedabyrene would be inhaled 3.6 times more rapidly w a C.E. for `JSP moth and Rowe 1975). me *de latitude h the results of C.E. calc~ations demonstrates the &pen&n= of the C.E. c.ahhtiOn On the numerical VahleS of the variables chosen, and correspondingly demonstrates the marked lotion of &e use of C.E. as an atmospheric measure of exposure b the wnb h en&m&d MOROCCO smoke. When the quantifica- con of an w ia needed, it is far more precise to use terms that defiae &e a of exposUre t0 the agent Of interest per unit he. However, the term cigarette equivalent is frequently used, not &ply 88 a mwure of exposure, but 88 a unit of disease risk that ~~them~ured~uresintoatikofdiseaseusingthe known daeresponse relationships between the number of ciga- re#es~~perdayandtheriskofdiseaee.IfC~.istobeusedasa tit of risk, the variables used to convert atmospheric measures into levels of rid for the active smoker need to be determined on the basis of the depcsition and smoke exposure measures for the average smoker. The deposition fraction of individual smoke constituents in t&e population of active smokers is needed rather than the range ob~rved in a few individuals. In addition, the actual average yield of the cigarettes smoked by the subjects in the prospective mortality studies would be needed to compare the dose-reeponse relationships accurately. The yield using the Federal Trade &nmission (Fl'c method may dramatically underestimate the actual yield of a cigarette when the puff volume, rate of draw, or number of puffs is increeeed; therefore, calculations using the Fl'C numbers may be inaccurate, particularly for the low-yield cigarettes. These limita- tions make exlxapolation from atmospheric measurea to c&are* equivalent units of disease risk a complex and potentially meanin- BleseP~. lAaiwa of Absorption In contrast, measuma of absorption of environmental tobacco smoke, particularly cotinine levels, can potentially overcome some of the limitations in translating environmental tobacco smoke expc+ sure3 into expected d&ease risk. Urinary cotinine levels are a reLatively accurate dosage measure of exposure to smoke; they have been measured in populations of smokers and nonsmokers, and are not subject to emrs in estimates of the minute ventilation or yield of the average cigarette. Potential differences in the half-life of cotinine in smokers and nonsmokers, differences in the absorption of nicotine relative to other toxic agents in the smoke, and differences in the ratio of nicotine to other toxic agents in mainstream smoke and sidestream smoke remain sources of error, but the accuracy with which active smoking and involuntary smoking exposure can be 200