TAYLOR, J.C., OEY, L.. MITTMAN, C. Protection of alpha,-protease inhibitor by plasma antioxidants: Potential abnormality in chronic obstructive pulmonary disease (COPD). Chest 83(5, Supplement): 905-925, 1983. THOMAS, E.D.. RAMBERG, R.E., SALE, G.E., SPARKES, R.S., GOLDE, D.W. Direct evidence for a bone marrow origin of the alveolar macrophage in man. Science 192(4243): 1016-1018, June 4,1976. THOMAS, W., HOLT, P.G., KEAST, D. Effect of cigarette smoking on primary and secondary humoral responses of mice. Nature 243(5404): 246-241, May 25, 1973. THOMAS, W.R., HOLT, P.G., KEAST, D. Development of alterations in the primary immune response of mice by exposure to fresh cigarette smoke. International Archives of Allergy and Applied Immunology 4W4): 481-486,1974a. THOMAS W.R., HOLT, P.G., KEAST, D. Humoral immune response of mice with long-term exposure to cigarette smoke. Archives of Environmental Health 3Ot2): 78-80, February 1975. THOMAS, W.R., HOLT, P.G., KEAST, D. Antibody production in mice chronically exposed to fresh cigarette smoke. Experientia 3Of12): 1469-1470, December 15, 1974b. THOMSON, M.L., SHORT, M.D. Mucociliary function in health, chronic obstructive airway disease, and asbestosis. Journal of Applied Physiology 266): 535-539, May 1969. TOIGO, A., IMARISIO, J.J., MURMALL, H., LEPPER, M.N. Clearance of large carbon particles from the human tracheobronchial tree. American Review of Respimtory Disease 87(4): 487492, April 1963. TOKORO, Y., EISEN, A.Z., JEFFREY, J.J. Characterization of collagenase from rat skin. Biochimica et Biophysics Acta 258(l): 289-302, January 20,1972. TOOMES, H., VOGT-MOYKOPF, I., HELLER, W.D., OSTERTAG, H. Measurement of mucociliary clearance in smokers and nonsmokers using a bronchoscopic video technical method. Lung 159(l): 27-34, March 1981. TRAVIS, J., BAUGH, R., GILES, P.J., JOHNSON, D., BOWER, J., REILLY, C.F. Human leukocyte elastase and cathepsin G: Isolation, characterization and interaction with plasma proteinase inhibitors. In: Haveman, H., Janoff, A. (Editors). Neutral P&eases of Human Polymorphonuclear Leukocytes. Baltimore, Urban Schworzenberg, 1978, pp. 118128. TSUCHIYA, M., KENSLER, C.J. The effects of autonomic drugs and 2-amino-2- methyl propanol, a choline antagonist, on ciliary movement. Federation Proceed- ings 18: 53,1959. TWUMASI, D.Y., LIENER, I.E. Proteases from purulent sputum: Purification and properties of the elastase and chymotyrpsin-like enzymes. Journal of Biological Chemistry 252: 1917-1926.1977. U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. A Report of the Surgeon General: 1971. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. DHEW Publication No. (HSM)71-7513,1971,458 pp. U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. A Report of the Surgeon General: 1972. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. DHEW Publication No. (HSM)72-7516,1972. U.S. PUBLIC HEALTH SERVICE. `The Health Consequences of Smoking. A Report of the Surgeon General; 1973. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. DHEW Publication No. (HSMj74-8704.249 pp. U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. A Report of the Surgeon General: 1974. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. DHEW Publication No. (CD074-8704, 1974, 124 pp. 326 U.S. PUBLIC HEALTH SERVICE. The Health Consequences of Smoking. A Report of the Surgeon General; 1975. U.S. Department of Health, Education, and Welfare, Public Health Service, Health Services and Mental Health Administration. DHEW Publication No. (CDC)76-8704.1975,235 pp. U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES. The Health Conse- quences of Smoking. The Changing Cigarette: A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHHS Publication No. (PHS)81-50156,1981,269 pp. U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE. Smoking and Health: A Report of the Surgeon General. U.S. Department of Health, Education, and Welfare, Public Health Service, Office of the Assistant Secretary for Health, Office on Smoking and Health. DHEW Publication No. (PHS)79-50066,1979. VALAND, S.B.. ACTON, J.D., MYRVIK, Q.N. Nitrogen dioxide inhibition of viral- induced resistance in alveolar monocytes. Archives of Environmental Health 20: 30%309,1970. VASSALLO, C.L., DOMM, B.M., POE, R.H., DUNCOMBE, M.L., GEE, J.B.L. NO, gas and NO,-effects on alveolar macrophage phagocytosis and metabolism. Archives of Environmental Health 26: 270-274,1973. VILLIGER, B., BROEKELMANN, T., KELLEY, D., HEYMACH, G.J., III. McDON- ALD, J.A. Bronchoalveolar fibronectin in smokers and nonsmokers. American Review of Respiratory Disease 124(5): 652-654, November 1981. VOSBRAT, L.C., RUMKE, P.H. Immunoglobuline concentraties, PHA reacties van lymfocyten in vitro en enkele antistof titers van gezonde rakers. [Immunoglobulin concentration, PHA reactions of lymphocytes in vitro, and certain antibody titers of healthy smokers.] Jaarboek Kenkerondenoek en Kankerbestrijding 19: 49-53, 1969. WANG, H. Different responses of paramecium aurelia to cigarette components. Nature 197: 946,1963. WANNER, A., HIRSCH, J.A., GREENLETCH, D.E., FORREST, T., SWENSON, E.W. Tracheal mucus velocity in beagles after chronic exposure to cigarette smoke. Archives of Environmental Health 27: 370-371,1973. WANNER, A. Clinical aspects of mucociliary transport. American Review of Respimtory Disease 116: 73-125,1977. WANNER, A. The role of mucociliary dysfunction in bronchial asthma. American Journal of Medicine 67(g): 477465, September 1979. WARR, G.A. The biology of normal human bronchoalveolar cells. Bulletin Europeen a'~ Physiopathologie Respimtoire 15(3): 23-34,1979. WARR, G.A., MARTIN, R.R. Chemotactic responsiveness of human alveolar macro- phages: Effects of cigarette smoking. Infection and Immunity 9(4): 769-771, April 1974. WARR, G.A., MARTIN, R.R. Immune receptors of human alveolar macrophages: Comparison between cigarette smokers and nonsmokers. Journal of the Reticu- loendothelial Society 22(3): 181-187, September 1977. WARR, G.A., MARTIN, R.R. Histochemical staining and in vitro spreading of human pulmonary alveolar macrophages: Variability with cigarette smoking status. Journal of the Reticuloendothelial Society 23(l): 5=2, January 1978. WARR, G.A., MARTIN, R.R., HOLLEMAN, C.L., CRISWELL, B.S. Classification of bronchial lymphocytes from smokers and nonsmokers. American Review of Respiratory Disease 113: 96-100,1976. WARR, G.A., MARTIN, R.R., SHARP, P.M., ROSSEN, R.D. Normal human bronchial immunoglobulins and proteins. Effects of cigarette smoking. American Review of Respiratory Disease 116: 2530,1977. WEISS, W., WEISS, W.A. Effect of tobacco smoke solutions on paramecium. Archives of Environmental Health 9: 500, 1964. 327 WEISSBECKER, L., CARPENTER, R.D., LUCHSINGER. R.C.. OSDENE, T.S. In vitro alveolar macrophage viability. Effect of gases. Archives of Environmental Health 18: 756-759,1969. WEISSBECKER, L., CRAMER, R.M., CARPENTER, R.D. Cigarette smoke and tracheal mucus transport rate. American Review of Respiratory Disease 104: 182- 187.1971. WEST, J.B. Distribution of mechanical stress in the lung, a possible factor in localization of pulmonary disease. Lancet 1: 83%&l, 1971. WHITE, R.R., JANOFF, A., GORDON, R., CAMPBELL, E.J. Evidence for in vivo internalization of human leukocyte elastase by alveolar macrophages. American Review of Respiratory Disease 125(6): 779-781, June 1982. WHITE, R., LEE, D., HABICHT, G.S., JANOFF, A. Secretion of alphal-proteinase inhibitor by cultured rat alveolar macrophages. American Review of Respiratory Disease 123(4): 447449, April 1981. WILDE, W. Mucociliary clearance of the slug mucosa exposed to dust and cigarette smoke. (Letter). Medical Journal of Australia l(9): 494-495, May 2, 1981. WOOD, R.E., WANNER, A., HIRSCH, J., DI SANT'AGNESE, A. Tracheal mucociliary transport in patients with cystic fibrosis and its stimulation by terbutaline. American Review of Respiratory Disease 111: 733-738,1975. WRIGHT, D.. GALLIN, J. Secretory responses of human neutrophils. Journal of Immunology 123: 285-293,1979. WRIGHT, R.R., STUART, C.M. Chronic bronchitis with emphysema: A pathological study of the bronchi. Medicinae Thoracalis 22: 210,1965. WYNDER, E.L., GOODMAN, D.A., HOFFMANN. D. Ciliatoxic components in cigarette smoke. II. Carboxylic acid and aldehydes. Cancer 18: 505-509, 1965a. WYNDER E.L., GOODMAN, M.S., HOFFMANN, D. Ciliatoxic components in cigarette smoke. III. In vitro comparison of different smoke components. Cancer 18: 1652-1658,1965b. WYNDER, E., KAISER, H.E., GOODMAN, D.A., HOFFMANN, D. A method for determining ciliostatic components in cigarette smoke. Cancer 16: 1222,1963. YAGER, J.A. ELLMANN, H., DULFANO, M.J. Human ciliary beat frequency at three levels of the tracheobronchial tree. American Reuiew of Respimtory Disease 121(4): 661-665, April 1980. YEAGER, H., Jr. Alveolar cells: Depressant effect of cigarette smoke on protein synthesis Proceedings of the Society for Experimental Biology and Medicine 131(l): 247-250, May 1969. YEATES, D.B., ASPIN, H., LEVISON, H., JONES, M.T., BRYAN, A.C. Mucociliary tracheal transport rates in man. Journal of Applied Physiology 39(3): 487495, September 1975. YORK. G.K., ARTH, C., STUMBO, J.A., CROSS, E.C., MUSTAFA, M.G. Pulmonary macrophage respiration as affected by cigarette smoke and tobacco extract. Archrues of Envrronmental Health 27(2): 9&98, August 1973. ZARKOWER, A., MORGES, W. Alteration in antibody response induced by carbon inhalation: A model system. Infection and Immunity 5(6): 915-920, June 1972. ZARKOWER, A., RHOADES, R.A., ROTHENBACHER, H.J. Effect of inhaled carbon on immune response in mice and rats. Proceedings of the Canadian Society of Biological Sciences 13: 641, 1970. ZWICKER, G.M., FILIPY, R.E., PARK, J.F., LOSCUTOFF, SM., RAGAN, H.A., STEVENS, D.L. Clinical and pathological effects of cigarette smoke exposure in beagle dogs. Archives of Pathology and Laboratory Medicine 102(12j: 623-628. December 1978. 328 CHAPTER 6. LOW YIELD CIGARETTES AND THEIR ROLE IN CHRONIC OBSTRUCTIVE LUNG DISEASE 329 480-144 0 - 85 - 12 CONTENTS Introduction Problems of Measurement by Machine Effect of Low Tar and Nicotine Cigarettes on Cough and Phlegm Production and Development of Chronic Obstructive Lung Disease Epidemiologic Studies Mechanisms of Lung Damage Variation in Smoking Pattern With Switching to Low Tar and Nicotine Cigarettes Smoking Behavior Carbon Monoxide Uptake Nicotine Uptake Role of Tar Content Variations in Pattern of Cigarette Smoke Inhalation Use of Additives in Low Tar and Nicotine Cigarettes Research Recommendations Summary and Conclusions References 331 Introduction Following the initial reports in the early 1950s linking cigarette smoke with lung cancer, the pathogenic role of cigarette tar content received considerable emphasis. Because the tar fraction of the smoke contained the bulk of the carcinogenic effect of whole smoke, and because lung cancer risk was closely related to other measures of total smoke exposure (number of cigarettes smoked per day, depth of inhalation, etc.), it was suggested that risk might be related to the amount of tar generated by different cigarettes. This prompted health authorities to advise smokers who were unable to quite smoking to switch to low tar cigarettes (U.S. Senate 1967; Health Department of the United Kingdom 1976). To facilitate this process, the Federal Trade Commission published smoking-machine assays of the tar and nicotine yield of different cigarette brands (Pillsbury et al. 1969). This approach to low tar and nicotine cigarettes was based on the assumption that smoking lower yielding brands, as deter- mined by a smoking-machine, would result in a proportional reduction in the lung's exposure to these toxic substances. This approach to "safer" cigarette smoking has been promoted by the tobacco industry and apparently accepted by the smoking public, as evidenced by the escalation in sales of low tar and nicotine cigarettes. However, there is increasing evidence that this concept of a "less hazardous" cigarette is misleading; although definitive studies are still awaited, it appears that switching from regular to low tar and nicotine cigarettes may not substantially reduce the risk of chronic airflow obstruction. Problems of Measurement by Machine The first step in evaluating the relative health risks of different cigarettes is to establish some standardized measure of the toxic substances in different cigarettes in order to facilitate comparison. Quantifying each of the several thousand constituents of cigarette smoke for each brand of cigarette, and assessing the changes in these constituents as the manufacturing and agricultural processes change, would be a truly herculean task; therefore, a more modest goal of quantifying tar and nicotine yields was accepted. To date, the yields determined by the Federal Trade Commission have been the most widely adopted. These measurements are obtained with a laboratory smoking-machine, which consists of a syringe pump that takes a 35 ml bell-shaped puff from a cigarette, over a 2-second period, once per minute until a predetermined butt length is reached, either 23 mm for nontiltered cigarettes or 3 mm longer than the filter overwrap for filter-tipped cigarettes (Pillsbury et al. 1969). These parameters are based on observations of smoking patterns in seven subjects in Europe in 1933 (Kozlowski 1983). Today's cigarette 333 is markedly different from that smoked in 1967 when these parameters were established, yet the same parameters are still employed. Measurements obtained using these parameters indicate a marked reduction in the tar and nicotine yield of cigarettes over the last decade (Figure 1). In addition to the actual tar and nicotine yield of the tobacco, the yield measured by a smoking-machine is influenced by many factors, including cigarette length and diameter, porosity of the cigarette paper, presence of a ventilated or an unventilated filter, butt length, number of puffs, interpuff interval, puff volume, puff duration, puff pressure profile, and frequency of puffing at different stages of cigarette consumption. The number of puffs is important in determining the tar yield of a cigarette, and the number of puffs taken from some brands with the official smoking- machine has significantly declined in recent years (Kozlowski 1981). Since puffs are taken at l-minute intervals, a more rapidly burning cigarette will have a smaller number of puffs. The burning time of the cigarette is determined by porosity of the cigarette paper, the- amount of tobacco in the cigarette, and the diameter of the cigarette column. In a survey of Canadian cigarettes between 1969 and 1974,- Kozlowski et al. (1980b) noted a significant reduction in the number- of puffs taken in the official assays over this time period, which was strongly correlated with a reduction in tar yield. Omission of the last- few puffs can markedly affect tar yield, because tar delivery- increases with each puff, and the last few puffs from a cigarette can contain twice as much tar as the first few puffs (Wiley and Wickham 1974). Currently published yields do not indicate the number of puffs taken, which may range from 7 to 12 and may result in a marked variation of the tar yield. Ventilated cigarette filters, which cause inhaled smoke to be diluted with air, are one of the major methods of achieving low tar yields (Gori and Lynch 1978). Cigarettes with ventilated filters constituted about 25 percent of all cigarette sales in the United States in 1979 (Hoffmann et al. 1980). During systematic interviews, Kozlowski et al. (1980a) found that from 32 to 69 percent of low tar smokers block these filter perforations with their fingers or lips, a feature unaccounted for by smoking-machines. This hole blocking increased the yield of toxic products by 59 to 293 percent. If a person smokes a cigarette in a manner identical to the smoking-machine, the delivery of tar and nicotine to the mouth will be the same as that estimated by the machine. Human smoking patterns are diverse, however, and show considerable variation from the machine parameters; puff volumes range from less than 20 ml to more than 90 ml (Tobin and Sackner 19821, compared with the fixed 35 ml volume employed by the machine. Differences in puff profile from the bell-shaped puff used by the machine also alter cigarette 334 Tar mg 100 -' 1950 1955 1960 1965 1970 1975 1980 FIGURE l.-U.S. sales-weighted average tar and nicotine yields SOCRCE Amencan Cancer Scaety 11981 I yield. Numerous studies indicate that smokers compensate for lower yielding cigarettes by altering their style of smoking. For each different cigarette brand, smokers may have a different smoking oattern. To provide more meaningful information, smoking-ma- :hines should be designed to reproduce variations in the manner of smoking each cigarette brand, and their assays should provide both m average and a range of tar and nicotine yields depending on the ndividual pattern of smoking (USDHHS 1981). Many investigators have examined the relationship between the nachine-determined nicotine yield of a cigarette and the concentra- ion of nicotine or its metabolites in blood or urine. A fair correlation vas observed in some studies (Goldfarb et al. 1976; Herning et al. 9831, but most studies have revealed a poor correlation (Russell et -1. 1975, 1980; Sutton 1982; Feyerabend 1982; Benowitz et al. 1983). Jachinedetermined nicotine yield accounts for only from 4 (Russell t al. 1980) to 25 percent (Herning et al. 1983) of the variation in lood nicotine concentration, whereas 50 to 60 percent of the ifferences in blood nicotine levels are attributable to individual smoking behavior. The overriding importance of the pattern of smoking in determining nicotine delivery from a cigarette was underlined in a recent study demonstrating that the nicotine content of the unburned tobacco was similar for cigarettes with high and low nicotine yields determined by smoking-machine assays (Benowitz et al. 1983). The concept of providing the smoker with information on cigarette yield need not be abandoned. Smoking-machines can be designed to control the puff number, puff volume, puff pressure profile, puff duration, puff interval, butt length, position of the cigarette during and between puffs, and "restricted" or "free" smoking, i.e., whether the butt end is closed or open Kreighton and Lewis (1978a, b). These parameters should be determined and used to obtain an average and a range of yields for each brand. Measurement of cigarette yield should include assays not only of tar and nicotine but also of carbon monoxide and other toxic substances, because compensatory smok- ing behavior may alter the exposure to each substance beyond that expected on the basis of tar and nicotine delivery. Effect of Low Tar and Nicotine Cigarettes on Cough and Phlegm Production and Development of Chronic Obstructive Lung Disease Cigarette smokers account for the vast majority of deaths from chronic obstructive lung disease (COLD) (Peto et al. 1983), and the relative risk for the effects of smoking on mortality from COLD is even greater than that for lung cancer (see the chapter on Mortality in this Report). Chronic obstructive lung disease in smokers may take the following three forms: (1) cough and mucus hypersecretion, (2) airway obstruction, and (3) emphysema. Frequently the three components coexist, as all are related to cigarette smoking, but the agents in cigarette smoke responsible for each type of lung injury may be different. Over the past 25 years, considerable progress has been made in our understanding of the role of cigarette smoking in the pathogenesis and natural history of COLD, but most of the available data have not related lung function to cigarette yield. Epidemiologic Studies The cardinal importance of cigarette smoking in the pathogenesis of COLD has been repeatedly documented, and generally the severity of disease increases with increasing cigarette consumption (Ferris et al. 1976). Because of this dose-response relationship, it has been hoped that a reduction in cigarette yield by filtration or other means would reduce the risk of disease (Gori 1976). Available epidemiologic studies of the effect of low yield cigarettes on the development of COLD have shown variable results, which reflects 336 marked differences between the studies in terms of the population studied, sample size, variation in cigarette brands, reference period of the study, criteria of respiratory involvement, and type of statistical analysis, and whether the study was of a cross-sectional or a longitudinal design. Separating the studies by the three compo- nents of smoking-induced COLD indicates that there is a growing body of data on the effect of cigarette yield on the development of mucus hypersecretion and airway obstruction, but currently no information on the development of emphysema. Several studies have examined the effect of cigarette yield on respiratory symptoms and have observed a relationship between reduction in cigarette yield and the prevalence of cough (Comstock et al. 1970: Freedman and Fletcher 1976; Fletcher et al. 1976; Dean et al. 1978; Schenker et al. 1982) and phlegm production (Comstock et al. 1970; Rimington 1972; Hawthorne and Fry 1978; Higenbottam et al. 1980b). Tar yield was not defined in some of these earlier studies (Comstock et al. 1970; Rimington 1972; Dean et al. 1978; Hawthorne and Fry 19781, but instead a comparison was made between smokers of plain cigarettes and smokers of filter-tipped cigarettes. The tar yield was specified in some studies: in the recent study by Schenker et al. (1982) it ranged from 0.4 to 28 mg; in the studies by Freedman and Fletcher (19761, from 17 to 20 mg; and in the studies by Higenbottam et al. (1980b1, from 18 to more than 33 mg, higher than that observed in many of today's cigarettes. In a cross-sectional survey of over 18,000 men (Higenbottam et al. 1980b), the beneficial effect of low tar cigarettes on phlegm production was lost when subjects smoked 20 or more cigarettes per day, as their prevalence of phlegm production increased to that observed in higher tar cigarette smokers. In contrast, in another cross-sectional study of 5,686 women (Schenker et al. 19821, cigarette tar content was a significant risk factor for chronic cough and of borderline significance for phlegm production; this effect of cigarette tar content was indepen- dent of the number of cigarettes smoked per day. Chronic cough or phlegm production was approximately twice as common in smokers of high tar (at least 20 mg) cigarettes as it was in low tar (less than 10 mg) smokers. In the latter study, however, multiple logistic regres- sion analysis indicated that the risk of chronic cough and phlegm production is more strongly affected by daily cigarette consumption than by tar content: these symptoms were 4.5 times more common in smokers of 25 or more cigarettes per day than in smokers of less than 15 cigarettes per day. A small number of studies have examined the importance of cigarette yield on change in pulmonary function. In a prospective study of 680 men, Comstock et al. (1970) noted that smokers of plain cigarettes, compared with smokers of filter-tipped cigarettes, had a lower FEV, at entry into the study. Followup measurements showed 337 I a23 24-27 2E32 .33 Tar per cigarette Prv$ FIGURE 2.-Relationship between mean FEVI of asymptomatic smokers (adjusted for height and weight) and tar yield of cigarettes, by number of cigarettes smoked per day SOURCE: Hlgenbottam et al rlS&Wbi. a greater mean reduction of FEVl in users of filter-tips, so that the reduction was similar in the two groups after 5 to 6 years of followup. Unfortunately, the variance of the data was not stated, and tests of statistical significance were not performed. In another longitudinal survey of 1,355 men, Sparrow et al. (1983) determined the effect of cigarette tar content, which ranged from less than 16 mg to more than 22 mg, on pulmonary function. Multiple regression analysis indicated that tar content did not significantly influence baseline spirometry or repeat measurements after 5 years of followup. Cross- sectional epidemiologic surveys also indicate no relationship be- tween abnormal pulmonary function and the use of filter-tipped versus plain cigarettes (Beck et al. 1981) or cigarette tar content (Higenbottam et al. 1980b) (Figure 2). Interpretation of these studies as evidence that cigarette tar and nicotine yield is not an important factor in the development of COLD is premature. First, cross-sectional studies are limited in their capability of defining the natural history of a disease. Second, COLD has a very slow progress, and Fletcher et al. (19761 suggest that a span of approximately 8 years is necessary to establish rates of change of spirometric values with sufficient confidence even to distinguish between smokers and nonsmokers. Third, we have no information on the baseline pulmonary function of smokers at the time they choose between high or low tar and nicotine cigarettes. Significant differences in pulmonary function have been observed between young adults who decide to smoke and those who avoid cigarette smoking (Tashkin et al. 19831, and it is possible that similar 338 function differences may exist in subjects who choose between high or low tar and nicotine cigarettes. Fourth, the yield of tar and nicotine used in many of these studies does not lie in the same range as that produced by many of today's cigarettes. However, the possibility that cigarette tar content is related to the development of cough and phlegm, but not of dyspnea or airflow obstruction, is consistent with current concepts of COLD. In a study of 792 men followed over an g-year period, Fletcher et al. (1976) observed that cigarette smokers were susceptible to two distinct chronic lung diseases-mucus hypersecretion and chronic airflow obstruction. This has recently been confirmed in a large prospective study (Peto et al. 1983) of 2,728 men, followed over 20 to 25 years, which showed that the risk of death from COLD was strongly correlated with initial degree of airflow obstruction, but bore no relationship to initial mucus hypersecretion. Given the evidence that mucus hypersecretion may depend on the tar fraction of cigarette smoke, while development of airflow obstruction is more closely linked to the ilumber of cigarettes smoked, Higenbottam et al. (1980b) reasoned that these differences might be due to a reduction in the particulate phase products, without a decrease in the gas phase products, in the low tar cigarettes. They hypothesized that tar droplets and soluble gases, such as sulfur dioxide and hydrogen cyanide, are more likely to be deposited or absorbed in the large airways where mucus is produced. The smaller airways, the earliest site of airflow obstruction, are exposed to a lower concentration of tar, but to a full concentration of insoluble gases such as nitrogen dioxide and ozone. This line of reasoning is in agreement with several studies showing a reduction in lung cancer with the use of low tar and nicotine cigarettes (Wynder et al. 1970; Lee and Garfinkel 1981; Rimington 1981; Hammond et al. 1976). The tar fraction is the component of cigarette smoke particularly linked with the development of both lung cancer and mucus hypersecretion. Although clinicians have long linked chronic bronchitis (mucus hypersecretion) with emphyse- ma, recent evidence indicates that mucus hypersecretion is not predictive of airflow obstruction, but is significantly greater in those smokers who develop lung cancer (Pete et al. 1983). Mechanisms of Lung Damage Studies of the mechanism of cigarette-smoke-induced lung damage have contributed significantly to the present understanding of COLD. Cigarette smoke may initiate and aggavate lung injury by a number of mechanisms and may also interfere with the lungs' defense responses. These mechanisms include the prot.ease-ir$$bit,or imbalance theo- ry for the pathogenesis of emphysema whereby alveolar wall 339 digestion results from an excess of proteases, a deficiency of their inhibitors, or a combination of both factors (see the chapter on Mechanisms in this Report). The sources of endogenous proteases include polymorphonuclear neutrophils and alveolar macrophages, both of which are found in increased number in the lungs of cigarette smokers. Protease release from both macrophages and neutrophils is increased in the presence of cigarette smoke (Rodriquez et al. 1977; Blue and Janoff 1978). In health, proteases are continually inhibited by al-antitrypsin, whereas proteases cause unimpeded digestion of lung tissue in patients with al-antitrypsin deficiency, with a markedly increased risk of emphysema. In addition to increasing the protease burden, cigarette smoke causes a functional inhibition of al- antitrypsin through the action of oxidants in cigarette smoke (Janoff et al. 19791. The relative potency of smoke from cigarettes of varying tar and nicotine yields in stimulating protease production and release and in inhibiting al-antitrypsin has received scant scientific investigation. Travis et al. (1980) tested the effect of both filtered and unfiltered cigarette smoke on the elastase inhibitory activity of al-antitrypsin. Filtered smoke reduced elastase inhibitory activity by 3 percent, and a 19 percent reduction was observed with unfiltered smoke; the tar content of the respective smokes was not stated. The researchers reasoned that this small in vitro effect would be greatly magnified by in vivo conditions in the lung, particularly through its huge surface area. In addition to examining the effect of filters, Cohen and James (1982) recently examined the effect of tar and nicotine content on the elastase inhibitory capacity of al-antitrypsin. The oxidant capacity of cigarette smoke was also examined using a chromogenic electron donor. Aqueous condensates of cigarette smoke were obtained from a variety of brands ranging in tar content from about 1 mg to more than 20 mg. Reported tar and nicotine content correlated well with the amount of measured oxidants and the ability of a brand to reduce the elastase inhibitory capacity of al-antitrypsin. Filters were found to remove 73 percent of the oxidants from the aqueous smoke solutions. While these findings suggest that low tar and nicotine or filter-tipped cigarettes could reduce a smoker's predisposition to enzymatic lung damage and consequent COLD, it should be noted that neither study examined the effect of lower yield cigarettes on protease production. Morosco and Gueringer (1979) demonstrated a greater increase in elastase in dogs exposed to high nicotine cigarette smoke compared with low nicotine cigarette smoke. More important, these studies have not taken into account the compensatory changes in smoking pattern likely to result with lower yield cigarettes. The airway response to acute exposure to cigarette smoke has been examined by several investigators employing spirometry (Da Silva and Hamosh 19811, body plethysmograph (Nadel and Comroe 1961j.I 340 and breathing pattern analysis (Tobin et al. 1982a). Airway narrow- ing has been consistently observed by some investigators (Nadel and Comroe 1961; Sterling 1967; Tobin et al. 1982a), but others report a variable response (Higenbottam et al. 1980a; Rees et al. 1982). In some studies, the acute airway response was unrelated to cigarette yield (Higenbottam et al. 1980a), but in most investigations (Robert- son et al. 1969; Tobin et al. 1982a; Rees et al. 1982), smoking a low tar or filter-tipped cigarette induced less acute bronchoconstriction. The acute airway response is probably localized to the larger airways, as acute cigarette exposure resulted in no change in the nitrogen washout test of small airway function (Da Silva and Hamosh 1973; Tobin et al. 1982al. These observations on the relative bronchocon- strictor response of various types of cigarettes may be important in our understanding of why some smoking novitiates persist with the habit despite the initial unpleasant reactions (Tashkin et al. 1983), but it is unlikely that repeated episodes of smoking-induced acute airway narrowing finally result in COLD. Future studies examining the mechanism of smoking-induced lung injury must not only take into account the range of cigarette yields, as determined by a smoking-machine, but also consider variations in smoking behavior. Puff volumes may vary considerably with nomin- al cigarette tar and nicotine content, thus altering the relative amount of various toxic substances yielded by different cigarettes. Similarly, inhalation profiles are of a diverse nature (Tobin et al. 1982b) and are likely to significantly alter the distribution, penetra- tion, and retention of cigarette smoke constituents in the lungs. Variation in Smoking Pattern With Switching to Low Tar and Nicotine Cigarettes Low tar and nicotine cigarettes have gained considerable populari- ty among the smoking public, partly on the premise that a reduction in the nominal tar and nicotine yield results in a proportional reduction in the health hazards of cigarette smoking. The validity of this approach to cigarette smoking is contingent on the accuracy of smoking-machines in reflecting the actual manner of puffing and also on the smoker not altering smoking behavior to compensate for variations in nominal tar and nicotine content. Should smokers develop compensatory alterations in their smoking behavior, this would not only reduce the relevance of the smoking-machine assays but might also alter the proportionate delivery of the different toxic substances in cigarette smoke and expose the smoker to concentra- tions beyond those predicted by the smoking-machine. 341 Smoking Behavior Nearly 40 years ago, Finnegan et al. (1945) studied the effect of alterations in cigarette nicotine content on smoking behavior and noted no change in cigarette consumption. It is only in the last decade, with the increasing popularity of low tar and nicotine cigarettes, however, that this question has attracted signEcant interest. The results of 38 studies examining alterations in smoking behavior with a reduction in cigarette yield are shown in Table 1. Considerable differences can be observed between the studies, partly reflecting variations in the level of cigarette yield reduction, alterations in other cigarette constituents, type and duration of switching procedure, parameters evaluated, and techniques used in their measurement. Most studies agree that smokers rarely increase their daily cigarette consumption upon switching from higher to lower yield brands. Reports are almost equally divided as to whether a smoker increases the number of puffs per cigarette or shows no change on switching to a lower yielding brand. There is an almost unanimous consensus that smokers take a larger puff volume from a lower yielding brand. Studies of puff volume also indicate huge variation between individual subjects (Guillerm and Radziszewski 1978; Hern- ing et al. 1981; Tobin and Sackner 1982; Herning et al. 1983) and that considerable increases in puff volume may occur on switching from a higher to a lower yielding brand, with certain subjects increasing their puff volume by up to 75 percent (Tobin and Sackner 19821. This compensatory increase in puff volume may be observed within a single experimental session (Tobin and Sackner 1982) and maintained over several weeks Rawbone et al. 1978; Stepney 1981). Full compensation for a lower yielding cigarette is generally not achieved by smokers taking a large puff volume (Rawbone et al. 1978; Herning et al. 1981; Tobin and Sackner 1982). Instrumentation is required to quantitatively assess the pattern of smoking, but it is important to realize that such instrumentation may, in itself, alter usual smoking behavior. Puff volume has been almost universally measured by using a specialized cigarette holder incorporating different flowmeter designs (Frith 1971; Adams 1977; Rawbone et al. 1978). These devices consist of two tubes connected to a pressure transducer that measures the pressure drop across a small resistance (a filter insert) in the holder; the flow measured is integrated to obtain volume. Use of a cigarette holder has been shown to increase the rate of puffing and puff volume, compared with measurements made with the cheek inductive plethysmogra- phy coil (Tobin and Sackner 1982). Unlike the compensatory increases in puff volume, measurements of the subsequent inhalation volume-which includes the volume of smoke mixed with air inhaled into the lung-have shown no change 342 TABLE l.-Effect of smoking low yield cigarettes on smoking pattern (`0 paramrters Nlcotme parameters ~____ Number Number Volume Durat Km Relationship Nlcotinr'rotlninr Mouth Relationship Reference Expenmental of cigs of Puff of of Expwd to level CCpOS"W LO First author year de@ smoked puffs/cig volume mhalatmn mhalatlon C'OHb (`0 nominal yield tblood!ur~ne/salw;i) Index normal yield F1llIIegaIl I945 Ashton 1970 Fnth 1971 (bhen 1971 Rwssell 1973 Turner 1974 Russell 1975 Coldfarb 1976 FO&8 I976 Adams 1977 Wald 1977 Sutton I978 R&s&one 1976 Schulz 197X (`wghtnn 1978a.b Adams 1978 (;udlerm 1978 clamk 1978 Ashton 1979 tiarfinkrl 1979 HI11 19fUl RobInson I980 Russell 1980 Hrnnmgtield 1980 Wald 19&I cl3 NC VCF CS CS cs t CS cs I CS f CS cs NC svs VCF NC C'S CS NC s CS cs CS CS NC svs NC cs svs svs VCF NC svs NC 1 /1N( NC NC I NC i N( t t 1 NC NC t t NC PcKJr (hod PO0 I 1 POW E TABLE I.-Continued Dr- CO paramrten Nicotme parameters NUdW NUdW Volume rhmitml Belatlonshlp Nwtmeicotinme Mouth Belatlonship Reference Experimental of CITY of Puff of of EXPlWd to level exposure t4, Fwst author year design smoked puffs/cig volume lnhaletwn lnhalatlon COHb (`0 nominal yleld ibloodod/urme/saliva~ Index normal yield Hernmg 1981 cs NC' I t Wtlld 19R1 svs NC I Poor Stepney 1981 cs NC t f I Fair I I Jaffe 1981 svs NC/ 1 NC PCXJr Tobin 1982 cs NC I NC NC Battig 1982 svs NC t NC Sutton 1982 svs t coed Poor Griffith8 1981 CS NC t Jaffe 1982 cs NC/ 1 NC Poor Feyrrabend 1982 svs Poor HtTWlg 1983 cs FHIT Benowitz 19n3 svs NOTE Cs controlled swtchmg. VCF - var,able c,garette filters. SVS a~wntaneo~rs voluntary wltchw. NC = no change. T - mcreaee. ; = dexeaee. CO - ctlrbon monvx,dr on switching to a low yield cigarette. Likewise, in one short-term study (Tobin and Sackner 1982), duration of inhalation showed no relationship to nominal cigarette yield. Perhaps compensatory changes in inhalation parameters require a longer period of time than puff volume does. Measurement of carboxyhemoglobin (COHb) concentration has been proposed as an index of the pattern of inhalation (Wald et al. 1975, 1978). While COHb provides valuable information on the amount of carbon monoxide absorbed from the lung during compen- satory alterations in smoking behavior, it is an indirect index and provides complementary information on cigarette smoke inhalation rather than replacing direct measurements of the volume of inhalation. Carbon Monoxide Uptake Unlike tar and nicotine, which are present in the particulate phase, carbon monoxide (CO) is a constituent of the vapor phase of cigarette smoke. For this reason, cigarettes purported to produce a low tar and nicotine yield may not necessarily provide a lower yield of carbon monoxide. Compared with tar and nicotine yield, carbon monoxide yield is more dependent on cigarette design, including such features as paper porosity and perforations in the filter tips. These factors regulate the dilution of smoke with air and the burning profile of the cigarette, and thus can significantly reduce carbon monoxide yield. Wald (1976) showed that the carbon monox- ide yield of filter-tipped cigarettes was 28 percent higher than that of plain cigarettes, although the average nicotine yield was lower in the filter-tipped cigarettes. He reasoned that smoke passing through a cigarette is diluted by air entering through the porous cigarette paper. However, the filter of filter-tipped cigarettes is surrounded by relatively nonporous paper, resulting in a higher content of carbon monoxide exiting from the proximal cigarette end. Perforations in the filter tip circumvent this problem and significantly reduce carbon monoxide yield (Hoffmann et al. 1980; Wald and Smith 1973). Many investigators have measured COHb or carbon monoxide concentration in expired gas following cigarette smoking and compared the levels achieved in smoking brands with different nominal yields (see Table 1). An increase, decrease, or no change in carbon monoxide intake has been observed, depending on relative differences in cigarette design and experimental procedure. As expected, unventilated filter-tipped cigarettes produced higher COHb levels than those observed with unfiltered cigarettes (Wald et al. 1977). This is in agreement with information provided by smoking-machine assays (Wald et al. 19731, but the use of ventilated filter-tipped cigarettes may produce COHb levels similar to those observed with unfiltered cigarettes despite lower carbon monoxide 345 yields on smoking-machine assay (Wald et al. 1977). Comparison of cigarettes with a marked difference in nominal carbon monoxide yield usually results in a lower COHb level when the lower yielding brand is being smoked (Russell et al. 1973; Turner et al. 1974; Sutton et al. 1978; Ashton et al. 1979); but over the range of different carbon monoxide yields there is a poor correlation between levels of COHb and measured carbon monoxide yield. Similar information has been observed using expired carbon monoxide concentrations. Nicotine Uptake It has been long considered that nicotine might serve as a primary reinforcer of cigarette smoking and that smokers might adjust their smoking behavior to regulate their level of nicotine intake. Several investigators have measured the blood, urinary, or salivary levels of nicotine or its major metabolite cotinine during the smoking of cigarettes of varying nominal nicotine yields (see Table 1). A reduction in blood (Russell et al. 1975; Sutton et al. 1978; Ashton et al. 1979; Hill and Marquardt 1980) and urinary (Goldfarb et al. 1976; Ashton et al. 1979; Stepney 1981) nicotine levels or in plasma (Hill and Marquardt 1980; Stepney et al. 1981) and urinary (Ashton et al. 1979; Hill and Marquardt 1980) cotinine levels has generally been observed on switching to a cigarette with a lower nominal nicotine yield. However, smokers show variable degrees of compensation for the lower yield, as there is generally a poor relationship between nominal nicotine yield and measured blood nicotine levels (Russell et al. 1980; Sutton et al. 1982; Feyerabend et al. 1982; Benowitz et al. 1983). Relating nominal nicotine yield and blood nicotine levels, Ashton et al. (1979) estimated that smokers compensated for about two- thirds of the difference in nominal yields when they switched from medium nicotine cigarettes to high or low nicotine brands. Using a stepwise multiple regression analysis of nicotine yield and blood nicotine concentration, Russell et al. (1980) observed a significant, but very weak, correlation (r = 0.21) between the two measurements, but the nominal nicotine yield of the cigarettes accounted for only 4.4 percent of the variability in blood nicotine concentrations. The use of absolute rather than logarithmic analysis in this study has been criticized (Kozlowski et al. 1982; Herning et al. 19831, and the criticism involved the problems of trying to predict doses to individuals rather than the dose to groups. In another study using log-linear regression analysis (Herning et al. 19831, a better correla- tion was observed between nominal nicotine yield and the increasing blood nicotine after smoking (r=0.5), but this study used Kentucky reference cigarettes rather than commercial brands, and these low yield cigarettes have less nicotine in the unburned tobacco than commercial low yield brands. Such a relationship still accounted for 346 only 25 percent of the individual differences in blood nicotine levels, whereas 50 to 60 percent was accounted for by individual differences in smoking behavior (Herning et al. 1983). Additional information on compensatory alterations in nicotine intake has been provided by studying the mouth exposure index, which is calculated from analysis of cigarette butts for nicotine content and a knowledge of the retention efficiency of the filter tip (Ashton and Watson 1970). Because the amount of nicotine retained by a filter is proportional to the amount that passes through, it is possible to estimate the amount of nicotine presented to the smoker from the nicotine content of the filter. Results using this index have revealed a greater variation between individual studies (see Table 1) than obsewed with blood nicotine measurements. This may be related to the fact that filter efficiency is usually determined by a machine, but retention of nicotine is also dependent on the way the cigarette is smoked; therefore, the retention efficiency of the filter may vary between smokers. Role of Tar Content The observations that smokers adapt their smoking behavior according to the nicotine delivery of a cigarette and that many of the toxic effects of smoking appear to be related to tar rather than nicotine content has led to the suggestion that altering the tar to nicotine ratio might produce a cigarette less hazardous to health (Russell 1976; Stepney 1981). A cigarette with a medium nicotine, low tar, and low carbon monoxide yield might be advantageous. While nicotine has been the component most extensively studied, it may not be the only substance responsible for the addictive power of tobacco. It is not possible to separate the effects of tar and nicotine in most studies, as their respective yields usually show a very close correlation. Using research cigarettes providing three different yields of nicotine and two different yields of tar, Goldfarb et al. (1976) found evidence of compensation for nicotine but not for tar content. The authors urged cautious interpretation of the results because of the limited range of tar yields examined. Examining a large number of subjects smoking cigarettes of varying tar and nicotine yield, Wald et al. (1981) found that both tar and nicotine were significantly related to blood COHb, taken as an index of cigarette smoke inhalation. Two- way analysis of variance of the data indicated that after allowing for the effect of either tar or nicotine yield, the COHb index was no longer significantly influenced by the other. A cross-over study of medium tar smokers who were switched to low nicotine, low tar cigarettes and medium nicotine, low tar cigarettes has been reported by Stepney (1981). While the intake of carbon monoxide was least with the medium nicotine, low tar cigarette, the mouth exposure 347 index to tar was similar among the brands. Indeed, the pattern of smoking adopted by the subjects was more effective in reducing the difference in tar delivery between the cigarettes than in compensat- ing for nicotine delivery. Further evidence indicating the importance of cigarette tar delivery in-determining smoking behavior was reported by Sutton et al. (1982). Using multiple regression analysis, they observed that when nicotine yield was controlled, smokers of lower tar cigarettes had higher blood nicotine levels than smokers of higher tar cigarettes, indicating that they inhaled a greater volume of smoke. In contrast, when tar yield was controlled, smokers of lower nicotine cigarettes had lower blood nicotine concentrations than smokers of higher nicotine cigarettes, indicating that they inhaled less smoke. These results suggest some compensation for tar over and above any compensation for nicotine. It may be that ncnpharmacologic, sensory stimulation by factors such as the flavor of cigarette smoke may be more important than nicotine in determining smoking behavior. These new observations, especially on the role of tar delivery, require further investigation. Most published research consists of controlled switching experiments in which the subject smokes cigarettes of varying yields (see Table 1). Further studies of smoking behavior in sutjects who have voluntarily chosen cigarettes of different yields are needed. The absence of an acceptable, palatable "standard" research cigarette continues to be an impediment to research in this area. Variations in Pattern of Cigarette Smoke Inhalation While cigarette smoking is the single most important factor in the development of COLD, the majority of smokers never develop clinically significant airflow obstruction (Fletcher et al. 1976). Despite the clear dose-response relationship between number of cigarettes smoked and death from COLD, attempts at identifying the indi-Jidual susceptible smoker on the basis of number of cigarettes smoked have had very limited success. Another approach to identifying the susceptible smoker is to study the manner of smoking, as this is probably a major determinant of the lung's exposure to cigarette smoke. Cigarette smoking consists of two phases: initially, the smoker takes a puff into the mouth, and after a variable 1 to 4 second pause, the smoke mixed with air is inhaled into the lungs (Rawbone et al. 1978; Higenbottam et al. 1980a; Tobin and Sackner 1982). Individual differences in the pattern of cigarette smoking such as the size of the puff volume, the duration of holding the smoke in the oral cavity before inhalation, and the depth and duration of inhalation are among the important factors determining the relative concentration of smoke constituents 348 that reach the lung. Despite its significance in determining the distribution and deposition of cigarette smoke, the mode of inhala- tion following the puff has received scant scientific investigation. A number of epidemiologic studies have examined the relationship between cigarette smoke inhalation, based on the smoker's subjec- tive estimation, and the severity of pulmonary disease. Results of these studies are conflicting; some investigators repcrted an associa- tion between smoke inhalation and the presence of mucus hyperse- cretion (Rimington 1974; Schenker et al. 1982; Dean et al. 1978) and decline in pulmonary function (Ferris et al. 1976; Bosse et al. 19?5), and others observed no relationship between inhalation and pulmo- nary dysfunction (Beck et al. 1981; Schenker et al. 1982). The inconsistencies in these epidemiologic studies may be due to the smokers' inability to accurately describe their inhalation pattern. There are three reports of the relationship between subjective estimations of cigarette smoke inhalation and direct objective measurement. Rawbone et al. (1978) found that the rating on a visual analog scale was a good predictor of inhalation volume (r=0.65). Conversely, Tobin et al. (1982a) noted no relationship between inhalation volume and the smoker's perception of depth of inhalation, indicated on a visual analog scale (i-=0.04); a similar finding was reported by Adams et al. (1983) (i-=0.04). Standardizing the inhaled volume for vital capacity did not improve the relation- ship. Other investigators using measurements of COHb observed a weak relationship between self-estimated inhalation and COHb concentration (Stepney 1982; Wald et al. 1978). Measurements of COHb reflect the amount of cigarette smoke absorbed by the long. In addition to being affected by the depth of inhalation, COHb concentration is influenced by the varying carbon monoxide yields of different cigarettes, the number of puffs per cigarette, puff volume, pulmonary function-particularly diffusing capacity and alveolar ventilation-and hemoglobin concentration (Wald et al. 1978; Ric- kert et al. 1980). Therefore, it yields valuable complementary information, but it does not provide a direct measure of the pattern of inhalation (Tobin et al. 1982a; Guyatt et al. 1983). Direct measurements of the pattern of cigarette smo!ce inhalation have been reported for a small number of smokers. Initially, the puff from the cigarette is taken into the molrth, and after a variable pause of 1 to 4 seconds, it is inhaled into the lungs (Rawbone et. al. 1978; Higenbottam et al. 1980a; Tobin and Sackner 1982; Tobin et al. 1982a; Adams et al. 1983). Higenbot,tam et a1,(1980a) reasoned that this pause, while holding the smoke in the mouth, minimized the irritant qualities of cigarette smoke. In a group of five subjects who were requested to inhale smoke directly into%eii lungs, withoslt an intervening pause in the mouth, consistent acute airway narrowing was observed. In contrast, smokers adopting the usual two-phase 349 smoking pattern showed a variable airway response. The authors suggested that buccal absorption of water-soluble compounds, such as sulfur dioxide and acrolein, together with precipitation of tar, minimized the irritating qualities of cigarette smoke. They observed no relationship between the acute airway response and amount of smoke inhaled in the regular two-phase smokers, although there appeared to be a relationship in those directly inhaling smoke into their lungs. However, there is a marked discrepancy in the inhala- tion volumes reported in this study compared with the values reported in other studies of cigarette smoke inhalation, probably due to the inaccuracy of the magnetometers employed for the measure- ments; therefore, a statement regarding the relationship between depth of smoke inhalation and the acute airway response may be misleading. The report that acute airway narrowing is uncommon after cigarette smoking is in disagreement with the findings of several investigators who have observed bronchoconstriction to be a common phenomenon after acute smoke exposure (Nadel and Comroe 1961; Sterling 1967; Da Silva and Hamosh 1981; Tobin et al. 1982a); however, it is certainly plausible that the response is greater in smokers who inhale smoke directly into the lungs than in two-phase smokers. The frequency of direct inhalation of cigarette smoke into the lungs is unknown. In a small study of 10 smokers, Tobin and Sackner (1982) observed 1 subject who showed an approximately 50 ml expansion of the abdominal compartment simultaneously with taking the puff from the cigarette. Adams et al. (1983) studied the relationship between puffing, cigarette smoke inhalation, and partitioning of airflow between the nose and mouth in 10 smokers. After taking the puff into the mouth, two subjects actively exhaled 80 ml and 200 ml volumes, respective- ly, before the subsequent inhalation. In this situation, the volumes of smoke might be expelled from the mouth, and little, if any, would be available for subsequent inhalation into the lungs. The frequency of this smoking pattern was not given, but another report from the same laboratory (Rawbone et al. 1978) indicated that it was uncommon. There was marked intersubject variation in the parti- tioning of airflow between the nose and mouth during smoking, with four subjects inhaling almost exclusively through the mouth, four inhaling predominantly through the nose, and the other two demonstrating both patterns of inhalation. The importance of factors in determining whether cigarette smoke is inhaled as a bolus followed by a subsequent "chaser" of air or is evenly distributed throughout the inhaled volume of air remains to be determined. Considerable discrepancies exist between published reports of the volume of air mixed with smoke that is inhaled into the lungs, with reported mean inhalation volumes of 34 to 152 ml (Higenbottam et 350