Tobacco Documents Online

BY. COPYRII~HT/~W OTfLE I~.U.S. Yields of Selected Toxic Agents in the Smoke ~f Canadian Cigarettes, 1969 and 1978. A Decade of ,Ckange?1' ~tVILLIAM S. RICKERTz AN~JAcK C. ROBINSON* .... :C~:.n~a~rKf:,~.~:w~ted~v6~,Vs-~f~e|ga~i¢ caf~on~ noxi-de (CO~' an0 tarhave evaluated for the decade ending 1979, Virtually no change occurred up to 1976; between 1976 and 1979 the sales-welghted average CO dropped from 21 to 16 mg with a much smaller decrease in the corresponding values for tar (16. I to 14.4 rag). Thus changes in CO deliveries did not take place at the same rate as changes in tar deliveries. If standardized smoking machine yields are related to the risk of smoking-related morbidity and mortality, the low correlation and the 8-year lag time in the curves describing the decrease in two of the major toxic constituents of cigarette smoke may be of imporlance in helping to explain a lower impact of less hazardous cigarettes on coronary heart disease (CHD) rates. Agsuming that e:~osure t~>~O is a risk factor for CHD, the reI,alively recent reduction in sales-welghted CO yields might have the effect of decreaslng CHD rates among smokers in the future. In a~ffi~gorr,to ~O yields,, HCN, ac.~oIei'n, a~d [o~al aldehyde yields of 2;-5 brands of cigarettes mar~fffas,~l.u~' in, 196~;' were comp.~iL~d, wgth yi:eld~ of the same cigarettes ma~mPactured in ' " l~g i~, ord~.o, assess ~hanges w, hieh may h~a~e.occurred in otl~r gas phase constituents. SIgn~can't decreases were noted in the yields of all constitutents other than CO: the average decrff~se per brand was HCN, 69 ~g; total aldehydes, 114/zg, and am:olein, _5,2 INTRODUCTION ']['he health consequences_, of smoking have bccn wc]] recognized _tot several -decadcsl particularly with respect to lung cancer, chronic bronchitis, emphy- :. sema, and coronary heart disease. Yet teenagers are still ta~. ing up the habit in large numbers and at least one out of three individuals in most Western countries continues to smoke (16). In recognition of this fact the U.S. National Cancer Institute and the U.S. Department of Agriculture jointly have supported an exten- sive program to produce and evaluate experimental cigarettes which are potentially less harmful to smokers who cannot or will not quit (6, 20). An important aspect of this "less hazardous" cigarette program has been the identification and quantitation of compounds in tobacco smoke which are biologi- cally active in various assay systems. Reports suggest that there are about 10 different constituents in this category which are present in concentrations ex- ceeding 1 /zg per cigarette; these include compounds which are toxic (carbon monoxide, nitrogen oxides, hydrogen cyanide, and nicotine), cilia toxic agents (acetaldehyde, formaldehyde, acrolein, hydrogen cyanide, phenol, and cresols), cocarcinogens (catechols and formaldehyde), and carcinogens and tiamor ini- Supported by Health and Welfare Canada grant for Research on Drug Abuse. To whom reprint requests should be addressed. 353 0091-7435~811030353-11502.00/0 CopyrJg.~t ~l 1981 b~" Ac~l.¢rnlc All rights of ~produetion in any form reserved. Tt05280027

354 RICKERT AND ROBINSON tiators (a complex group of compounds associated primarily with cigarette "'tar") (21). Because of the demonstrable biological activity of these compounds and in particular the so-called "'tar" fraction, changes in standardized cigarette deliv- eries "have been used as a potential indicator of the change in relative toxicity of cigarettes with time (1~6~ 20). ~aee 1950, ~en the first reports a~peared linking cigarette smoking to ~ng ~.1~4~ ~,, 2~, ~_gz~x~tte tar d~liw~ri~s~ have st~adi|y d~el~n~d ~sl~l~i~ i~ a measurable decrease in excess mortality for smokers who switched to "low" tar brands (7). With respect to specific diseases, the adjusted number of lung cancer ~w tad.u[~otin:~ smokers ra~ed from 57 to .8~ ~f the ad_j.~ted .~a~raber 6ff~-C~ ~s tr~f Ieg~ d~am-ati~:; the Val'ues ranged fromSl t093% for low tar[nicotine smokers in comparison with those who smoked the high tar/nicotine cigarettes (8). One of the more obvious reasons for the difference is simply that lung cancer rarely occurs in nonsmokers without a special occupational risk factor, unlike sudden death from CHD which is quite common in older nonsmokers and is associated with other factors such as hypertension, high-density lipoproteins, body weight, and family history. Since there are multiple risk factors for CHD, the eontribu,tioa of"low" tar cigarette usage a,lone, to the reportedr reduction in CFID mortality is difficult to assess particularly since CHD mortality rates have been d~re~":ng ~':oag nonsmokers. Also, there is the additional complication ,~hat s~ro"~]~r~0f.~i:~ ~#fes~. ha¢,a~do:u~'" eig-arett¢s may be si,mp;ly nro~e health eon, scious in, gene'fa[, modifying other risk factors when they change ~heir smoking habits. 8owever, the fact remains that the use of cigarettes is the most. potent con- tributor to CHD mo~tality (16). Thus, another hypothesis to account for the re- duced impact of"low" tar cigarettes on mortality rates for CHD is simply that the agent(s) which influence the initiation and/or progression of CHD were not re- duced significantly in_this-type of cigarette. For example, carbon monoxide has been identified in the 1979 report of the U.S. -Surgeon General as a possible critical factorin a number of conditions including coronary heart disease, athero- sclerosis and sudden death, chronic respiratory disease, and fetal growth retarda- tion (16). The resulting concern generated by this and other reports prompted a recommendation by the World Health Expert Committee on Smoking Control to enact legislation requiring a statement to appear on packets of cigarettes giving the average CO as well as tar and nicotine yields (2). In a recent study, we examined the gas phase yields of all commercially avail- able Canadian cigarettes and found that tar yield is a poor predictor of gas phase deliveries (12). The present investigation was undertaken to compare changes in sales-weighted brand yields of carbon monoxide and tar for an I l-year period (1969-1979) and to compare the gas phase deliveries of cigarettes purchased in 1969 with those acquired in 1978. Little change in the yields of gas phase con- stituents such as CO could be interpreted as supporting the hypo.thesis that expo- sure to cigarette smoke CO is an important risk factor in coronary heart disease. Cigarett Food and sampling ! randomly graphic re: stored at ~ sample wa and the re in 1'978 (3) cities acre Protection selected fr, 22°C for a length of 3 than 27 mr years and .~ States, wht Sales-we obtained b ( 1973-197.t piled by a I the Canadi~ (1968-_197~ lation of Canadian "I Before I! ~icotine wi' to establist to 1972. cou resulting re to the data i 0.998. The with which Analytica using a sam and Bird 2C ~ Although t • five cigarettes Thus a samplt Ti05280028

.e "tar") Is and in te deliv- • xicity of CIGARETTE YIELDS, A DECADE OF CHANGE? 355 MATERIALS AND METHODS C~garette samples. Cigarettes were purchased on the open market by Canadian. Food' and' Drug Directorate Field' Inspectors during the first quarter of 1969. The sampling pattern consisted., of two packages of each brand purchased by two ,ra,~:.d~:m~ly selected .offices of the Food and Drug Directorate in each of five geo- :nicotine ettes (8). g cancer r, unlike "s and is proteins, ." I-ID, the t i~ CHD ~ve been tion that o/t,scious ~ habits. .ent con- ,r the re- ,~ that the .- not re- :xide has .possible , athero- retarda- ~mpted a ontrol to iving the ily avail- as phase ianges in tr period hased in ase con- lat expo- :lisease. Protection Branch, Health and Welfare Canada. A representative subsample was selected from the national sample and conditioned at 60%. relative humidity and 22°C for a minimum of 48 hr before testing. All cigarettes were smoked to a butt length of 30 mm or overwrap plus 3 mm when the overwrap length was greater than 27 ram. In Canada, the standard of 30 mm has been used for more than 10 years and yields results which are quite different from those obtained in the United States, where the standard is. 23 mm (see Ref. (12) for a comparison). Sales-welghted averages. Sales-weighted: averages for Canadian cigarettes were O~_~.~,ne:d' by "comMn~ng estimates of wet t~r (1968-197~) or dry tar deliveries (i~ 1~7~) with, s-al~s fig~utes provided by ~e,alt'h a~d: :~elfo~e Canada as com- ~dl~'y a ~yi~ate research organization (12'). These estimates were produced for th~ Canadia~ government by the Department of Statistics, University of Waterloo (1968-1976), and by Labstat Incorporated (1978-1979). Data for the calcu- lation of the Canadian sales-weighted averages for 1977 were provided by the. Canadian Tobacco Manufacturers Council, Before 1973, tar ~f.alues__were reported as TPM (total particulate matter) minus. nicotine with no correction for moisture content. Data obtained in 1973 were used to establish a relationship between wet and dry tar so that the results from 1968 to 1972 could be adjusted for the purpose of comparison. The very close fit of the resulting regression equation dry tar = 0~318 + 0.932 wet tar (30 mm) (30 ram) to the data is illustrated by a significance level well below 0.1% and anR" value of 0.998. The standard error of regression of ±0.024 mg demonstrates the accuracy with which mean dry tar can be predicted from mean wet tar. Analytical procedures. Total particulate matter (TPM) yields were determined using a sample of 50 cigarettes per branda half of which were smoked on a Phipps and Bird 20 port smoking machine under standard conditions (I 1) and the other a Although the current practice is to report yields on a per cigarette basis it should be pointed out that tlve cigarettes are required for each analytical value and the per cigarette yield is obtained by division. Thus a sample size of 50 cigarettes in fact yields only 10 independent observations. Tt05280029

356 RICKF-,RT AND ROBINSON half using a similar piston type 20 po~ instrument manufactured by Filtrona Inter- national. :Nicotine and water were estimated using a gas chromatographic proce- ,~Iu're, and "'tar" delive~es obtained by subtracting nicotine and water from TPM (17). Estimates of CO deliveries were obtained using a nondispersive infrared a, aaly~er (ATCOM), linked directly to a Model 300 Filtmna smoking machine. fl~'~Yetty connected to a gas chromatograph (14), The determination of HCN and acrolein yields of the 1969 cigarettes required the modi;fication of the Phipps and: Bird: 20 port smoking machine to accommodate l~l~'e ~ri~u.m~nl~hti, on-anc~ eh~mistri'~:~ n~ecessary for simul- taneous determination of all three analytes have been published previously (13). The actual values for acrolein, total aldehydes, and HCN deliveries of the 1969 cigarettes used in this study are the averages of 25 cigarettes (five determinations) per brand. HCN yields are the sum of the gas phase deliveries and the HCN content of TPM estimated using aliquots of the solution prepared for nicotine and water determinations. Brand by brand listings for all five gas phase constituents in 1978 cigarettes have been pub}ished previously (12). RESULTS ,~al~W<eighted Tar and ~0 Deliveries I~n Canada a-s in the rest of the world, sales-weighted tar averages have dropped considerably dudngthe past 10 years, from 21.1 mg in 1968 to 14.4mg in I979 for a change of about 32% (Fig. 1). A similar decrease in sales-weighted average yields I~ I I I l I l I I I I I Fro, I. Canadian sales-weighted cigarette tar, O: and cigarene CO deliveries. 121: for the period 1968-19"/9. The first arrow indicates the point al which official government concern was expressed with respect to CO deliveries and ~he second arrow marks the official Canadian Government release of a brand by brand listing of CO yields. of CO that as CO yi~ in a di! cries. ! son o~f to a di Since t mos~ b bear rr "low-y is patti vented the sin- Ref. ( I leveled weight~ largely brands the ch~ of filte. weighh which ~ _ ILls; lhe bra 1969~h tar, L~ yields ~ ofthel ~'low-y "Arbit n Data

ona Inter- hic proce- :rein TPM _-. infrared machine. ' L.smok~g ~uired the ~mmodate msly (1 f the 1969 ninations) the HCN :otine and tituents in e ~roi~ped 1:9~9 for a age yields CIGARE-t'-TE YIELDS, A DECADE OF CHANGE? 357 of CO did not occur and, in fact, may have increased between 1968 and 1972. Be that as it may, for the period ending 1976, there was little change in sales-weighted :CO yieJ'ds with the v~lues rema,~ning c~nstant at approximate'ly 20,7 rag, resulting in a difference of 4.8 mg between sales-weighted tar and sales-weighted CO deliv- eries~ By m, id~I979, this difference was re.duced to only 1.6 mg. A direct compari- son ,of~ese values with, similar restflts fro~ ~t~he Ut~,ited' Stetes is trot. possible due ~o' a" fl~"/~e'rcnce ~fft tl~-e cffoi~ee df~ar~ a~'t~rO~t~a'te 'burr length for testing purposes. Since the U.S. standard is 23 ram, the Canadian results at 30 mm will be lower for most brands (see Methods). ~n#~-.~tmm~i.~ ~, -,~.~ "s~da~" _i,n: ira-'~ar~et-s~a~e from/T to 2~% .for "'low-yield" cigarettes ~ 12 rag) which took place between 1975 and 1978 (Table 1) is partly responsible for the decrease. These cigarettes are often constructed with vented filters which effectively reduce both tar and CO deliveries as illustrated by the simple correlation coefficients recorded in Table 2 (based on the data set in ReE (12). However, the market share for this type of cigarette appears to have leveled at about 20% in 1978 (Table 1) so that the 2.6 mg decrease in sales- weighted average CO yield noted between 1978 and 1979 must have been due ta~rge]y to a modification, ofexisting products rather than the introduction of new br_a,nds or a radica!; shift in smokers' pgeference. This suggestion iS ,supported by the ¢:han,ges w,'h[eh have oeeurr~d b~tw6¢n 1976 and 1979 in the IO Madi.ng hrartds 'o~ f/l~er c~g~eettc~-~opresenting o~ze~" 5~ o.f sa~l~s. During this period the sales- weighted average yield of CO for this group dropped by 4.0 mg or an amount which is only 20% less than the drop recorded for all cigarettes (see Table 3). It is also interesting to note that the difference in the sales-weighted average for the brands in Table 3 and the "all-brand" average increases from year to year, In 1969 this difference was CO, 0.? mg; tar, 0.6 rag;: and had tripled to CO, 1.9 rag; tar, 1.9 mg by 1979. Obviously if a decrease ii~ s.ales-weighted smoking machine yields.is desirable, the most expedient way to bring it about is to modify those 10% of the producls which make up 50% of the market in addition to introducing new "low-yield" brands which may or may not prove acceptable to the consumer. r the period ts expressed nt release of TABLE i MARKET SHARE OF CANADIAN LOW TAR BRANDS. (1973--1979y' Market shard' Year (%) 1973 8.3 1974 6.5 1975 7.4 1976 12.5 1977 I6.4 1978 21.0 1979 20.0 " Arbitrarily defined as those brands whose nominal dcliver/es are less than or equal to 12 rag. ~" Data provided by the Canadian Tobacco Manufacturers Council. -1105280031

358 RICKERT AND ROBINSON TABLE 2 SIMPLE CORRELATION COEFFICIENTS FOR BRANDS OF CANADIAN CIGARETTES WITH T~R D~LIVER~S LESS THAN OR EQUAl TO 13 Hydrogen Carbon Total Nicotine cyanide monoxide aldehydes Acrolein TaT O. 92 0.91 0;90 0~89 ~N~gEne '0L86 0~Sq 0~79 Hydrogen cyanide 0.91 0.92 Carbon 0.96 "Thirty brands representing 19rA of total sales. All correlations are highly significant (P << 0.001). This point is well illustrated by Fig. 2 which demonstrates that the availability of low-yield ~ 12 mg) brands continued to increase through 1979, representing about 26% of the available brands at that time, although the preference for such brands reached a maximum of 21% in 1978 and dropped to 20% in 1979 as shown in Tabl~ 1. Clga, re~ Gas Phase Deliveries, 1969' and I978 ~'~W-.fi~e bra~ds w~re s~,l~d: for ~his sl~udy, r~presenting 84% of sales in 1969. The brand-by-brand results summarized in Table 4 indicate a striking differ- ence between changes jn CO yields and changes in the other gas phase con- stituents. With respect to CO; a total of 16 or64% of the brands tested either had an increased yield or no change; a decrease was noted for the remainder so that, overall,_ there was no net. change. However, the yields of other gas phase con- stituents did decrease significantly during the decade ending in 1978; the average .change with respect to HCN yields was 69/zg per_brand (29%), aldehydes, l 14 p.g per brand (11%), and aerolein, 5.2 ~g per brand (7%). DISCUSSION Cigarette carbon monoxide has been identified in the report of the U.S. Surgeon General as a possible critical factor in coronary heart disease and it is widely recognized that exposure to cigarette tar is an important factor for various types of cancer (16). Consequently, switching to cigarettes with reduced chemical yields should result in a lowering of the risks associated with smoking for those who do not change their smoking habits. With respect to lung pathology, the evidence suggests that there may be some benefit from switching to brands with low chemi- cal yields (1, 7, 23, 24), but with respect to coronary heart disease the results are not as encouraging. If the current debate with respect to cigarette carbon monoxide and coronary heart disease is to be resolved, a knowledge of changes in sales- weighted average CO deliverie's over time are important; particularly given the extremely poor correlation (r = 0.16, nonsignificant) between cigarette tar and CO for brands whose deliveries exceed 13 mg (about 80% of Canadian sales, see Table T105280032

0.96 .< 0.001). .bility of lg about brands nown in ~a!~s ,ie ,g dl~ffe r- :se con- ~her had - so that, tse con- average • 114 Surgeon , widely types of al yields who do .videnee ; chemi- suits are onoxide n sales- iven the and CO ~e Table CIGARE'I-FE YIELDS, A DECADE OF CHANGE'?. 359 TI05280033

360 RICKERT AND MONOXIDE DRY TAR o ~ 3O 40 0 8 16 24 0 8 24 0 8 IG 24 1969 /979 o e ~s.z.~ o -e CIGARETTE YIELD (rag) . FIG. 2. Histograms of the availability of cigarette brands in lar and CO segments of 4 mg for the years 1969, 1~5; and 1979. 5). The implication of this finding can be seen when predicting CO for a cigarette with a tar_ value of 15 rag; the resulting 95% confidence interval for CO is approxi- mately 15 to 23 mg and this interval includes 80 of 98 brands investigated (25). However, with respect to brands with tar deliveries ~< 13 mg, the correlation of tar and CO yields is quite high (r = 0.90; P << 0.001) presumably due to the incor- poration of air dilution devices into the design of most low-yield cigarettes (10). Based on the data presented here, it is more than likely that there was not a significant change in CO exposure for those individuals who were part of the extensive American Cancer Society study of the mortality rates of over one mil- lion smokers in relation to the tar and nicotine yields of their cigarettes (7, 8). Of course, this statement assumes little difference in the characteristics of cigarettes and smoker preference between Canada and the United States. In Canada, a radical change in CO yields did not take place until 8 years after the beginning of the rapid decline in tar deliveries; thus, the anticipated health benefits from such a change may not be evident for some time to come. Undoubtedly a smoker's concept of the risk associated with-his habit has played an important role in the recent popularity of lower-yield brands. However, TI05280034

51 62 66 72 66 74 83 68 70 52 60 81 65 59 62 -76 ' i 69 75 - 181j, 66 TA B'.,LE A COMPARISON OF THE GAS PH^~E PURCHASED IN 1969 CO (rag) HCN (~g3 Aldehydes (gg~~ Bmnds ]978 [969 At' 1978 1969 A 1978 1969 Belvedere KSF 21.2 19.0 Belvedere RF 18.3 15.7 Benson & Hedges 100's 21.7 21.8 Benson & Hedges 100's M I8.9 21.2 Cameo KSFM25.7 22.4 Craven A RF 14.9 12.5 Craven A KSF 17.9 23. l Craven M KSFM 15.6 16.7 DuMaudcr RF 20. I 17.5 DuMaurier KSF 21.9 26. I DuMaurier KSFM 17,9 21.5 Export A RF 16.8 18,7 Export A KSF 23.8 23.5 Export A RP 16.7 18.2 Mark Ten RF 2J.7 19.6 Matinee RF13.7 13.1 Matinee KSF 17.0 15.5 Number 7 KSF 22.5 23.1 Peter Jackson KSF 25.8 23.4 Players RP18. I 18.0 Players RF18.9 16.7 Players KSF 22.9 20.8 Rothmans KSF 21,I 22.0 Sportsman RP 15.8 17.8 Sweet Cap RP 16.2 17.5 1.2 2.6 185 215 - 30 1131 941 168 179 NS 864 877 NSr 150 272 - 22 1019 1368 -,2.3 ' + 3.3 +2.4 -5.2 -I.I +2.6 ' .-4.2 -3.6 - 1.9 ~ NS +2.1 NS +1.5 N$ +2.4 NS +2,2 +2,1 -2.0 -1.3 175 '223 II1 131 127 200 221 147 1190 197 161 169 !oo 13o ~87 233 ~76 196 218 166 17o 307 - ~3,2 1071 1398 284 "-6T 1108 1160 130 NS 654 875 265 - 134 946 1184 184 - 57 971 948 224 -24, 891 1048 304 -,83 981 1231 298 - 151 889 1371 228 -~8 909 929 292 -95 I161 1205 260 - 99 843 918 248 - 79 894 986 i24 - 24 810 748 178 - 48 825 1073 281 -94 1016 1177 304 - 71 1091 1159 229 - 53 908 I023 221 - 25 791 991 290 -- 72 1091 1240 261 - 95 ll41 1217 229 - 59 897 865 218 - 50 909 1090 1969 A 65 NS 62 NS 82 - 16 86 - 17 77 - 10 56 , NS 81 - 19 62 NS 69 NS 87 - 9 84 - 18 72 NS 85 NS 72 NS 70 NS 62 -10 64 NS 86 NS 77 - 12 69 - 10 65 - 3 76 - 7 80 NS 64 NS 69 NS "These brands represented approximately 84% of all sales in 1969 and 74r,6 in 1978. i ~' Difference: delivery for 1978 - delivery for 1969. : " Not significant: in order for the difference to be consldered significant (P ~ 0.025. two-tailed) the absolt~¢ y~lue must exceed 1.0 mg (CO), 19 ~g (HCN), and 116 #g faldehydes) based on the standard errors (see Ref. (121). ! o >

362 RICKERT AND ROBINSON TABLE 5 SIMPLE CORRELATION COEFFICIEh~'S, BRANDS OF" CANADIAN CIGARETT~ WITH TAR DEUVERIES GREATER THAN 13 Hydrogen Carbon Total Nicotine cyanide monoxide aldehydes Acrolein Tar '0.6"2' " 0~.3 8"" 0~ 16 IL28~' 0.24 ,blf~oti, n:¢ 0;4~** 0~0~ 0.23 0~0~* Hydrogen cyanide 0.68** 0.SP* 0.44** Carbon ,r~o~i,d¢ 0~67~'* 0~.32** Seventy-two brands representing 81% of total sales. * P ~< ** P~< 0.005. as. many as 80% of Canadian smokers currently smoke brands whose tar yield is in excess of 12 mg, making it likely that factors other than perceived hazard dominant in smokers' choice of brand (15, 18). Consequently there must be addi- tio~a~.,explanafions for the m.a, rked decrease in sales-weighted CO yields which o:¢curred b~tw~e~n, 197.6 and 19¢/9. In t976, concern was expressed officially by the Department of Health and Welfare Canad~ to the Canadian Tobacco Manufacturers about the CO yield of qigarettes and in 1979 the Minister of Health released the first brand by brand listing of the CO yields of all Canadian cigarettes (19). It is unlikely that the decrease in sales-~weighted average CO yield of 23% which occurred during this time inter~val is unrelated t9 these_two events. Evidently the Canadian tobacco manufacturing industry has substantially reduced cigarette y~elds in response to public demands and the release of brand by brand yields by the Canadian govern- ment as recommended by the World Health Organization's Expert Committee on Smoking Control (2). ACKNOWLEDGMENTS We are grateful for the competent technical assistance ofJ. Zavitsky, V. Patel, K. Sehulthies and S. Gibbons. REFERENCES 1. Auerbach. O.. Hammond, E. C., and Garlinkel, L. Changes in bronchial epithelium in relation to cigarette smoking. 1955-1960 U.S.; 1970-1977. New Engl. J. Med. 300. 381-386 ¢1979). 2. *'Controlling the Smoking Epidemic." Report of the WHO Expert Committee on Smoking Con- trol. Technical Series Report 363, pp. 57-72. WHO, Geneva, 1979, 3. Cold storage cools critics, keeps tax heat on cigarette industry. Tobacco Intern No. 30. 19-22 (1979). 4. Doll, R., and Hill, A. 13, Lung cancer and qther causes of death in relation to smoking. A second report on the mortality of British doctors. Brit. Med. J. 2, 1071-1081 (1956). 5. Forbes, W. F., Robinson, J. C., and Stanton, M. Tar and nicotine retrieval from cigarettes avail- able in Canada. Cancer 2~, 910-912 (1969). 6. f 7. } 8. I- 9. I 10. F 12. ~ 13. I~ 14. I~ 15. I~ 16.'" 17. S 18. S 19~-" " 20. q 2L ~. 22. % 23. X 24. X 25. ~. TI0528003~

eld is azard addt- ~: ~n-d .~ld of ~rand tt the g this ~acco -- lse tO vern- ee on and S. :tion to "9). g Con- 19-22 .econd , avail- CIGARETTE YIELDS, A DECADE OF CHANGE? 363 6. Gori, G. B., and Lynch, C. J. Toward less hazardous cigarettes: Currant advances. JAMA 240, 1255-1259 (19781. 7. Hammond, G. C., Garfinkel, L., Seidman, H., and Lew, E.A. Tar and nicotine content of cigarette sraolce in retation t~o death rates. Environ. Res. 12, 263-274 (1976). 8. Hammond, E. C...Garfinkel~ L., Seidman, H.. and Lew, E. A. Some recent findings concerning cigarette smoking, in "'Or~.in£ of Human,Cancer,'" Book A, "'Incidence of Cancer in Humans'" ,(J~. i~ Watson~and: ~. A. WInsten., Ed~.), Co.tddSpring Harbor Cont'erenees oa Cell Proliferation, 9. Levin, M. L., Goldstein, H., and Gerhardt, P. R. Cancer and tobacco smoking..IAMA 143, 336-338 (1950). I0. Parker, J'. A., and~ Momgomery, R. T. Desig~ criteria for ventilated filters. Beitr. TabaL~rsch. 10, l~':K~'~Cer~;:W, g.~Rob~nson, ~'~-C., anti Young~ J. C. Estimating the hazards of less hazardous cigarettes. I. Tar, nicotine, carbon monoxide, acrolein, hydrogen cyanide and total aldehyde deliveries of Canadian cigarettes. J. To.rfi.oL Environ. Health 6, 351-365 (19~). 13. Rickert, W. S.. and St~kweII, P. B. Automated determination of hydrogen cyanide, acrolein and total aldehydes in gas phase of tobacco smoke. J. Auto. O~em. 1, 152-154 (1979), 14. Robinson, J. C., and Forbes, W. F. The role of carbon monoxide in cigarette smoking. Arch. Environ. Health 31,425-434 (1975). 15. Rus~II, M. A. H., Harris, ~., and lyer, R. Relation of nicotine yield of cigarettes to blood nicotine level Of smokers Brit. Med. J., in p~ss (]980), 16, "~;king and Health," A Report of the Surgeon Oene~l (19~). DHEW Publication No. (PHS) 79-S~6. U.$. Department of Health, Education and Welf~. 1'. 81~an~ ¢. H., and SumMer, B:~ ~, Moisture ¢~t:~nt o~the paniculate phase of smoke from filter and 1'8. Stepn~y,, ~. Consffmption of,~garettes with reduced tar and nicotine delivery. Brit. J. Addk'tion 75, 81-88 19. "'Tar, Nicotine, Carbon Monoxide Figures Released.'" ~ess release by theHonourable David Crombiet Minigter of Health,. Health and Welfare, Canada, October 18, 1979. 20. Tso, T. C., and Chaplin, J. F. "'Simple Co,elation and Multiple Regression among Leaf Cha~c- re,sties, Smoke Components and Biological Responses of Bright Tobaccos.'" Technical Bulle- tin No~_1551, U.S. Depaament of Agriculture, 1971.- 21. Wynder, E. L., and Hoffman, D. Tobacco and health. A societal ch~lenge, NOr E¢igl. J, Med.. 300, 894-~3 22. Wynder, E. L., and Graham, E.A. Tobacco smoking as a possible-etiologic factor in bron- chiogenie carcinoma. JAMA 143, 329-336 (1950). 23. Wynder, E. L., Mabuchi, K., and Beattie. E. J. The epidemiology of lung cancer. JAMA 213, 2221-2228 (1970). 24. Wynder, E. L., and Stellman, S. D. Impact of long-term cigarette usage on lung and larynx cancer risk: A case-control study. U.S. Nat. Cancer lest. J. 62, 471-477 0979). 25. Young, L C,, Robinson, J. C., and Ricke~ W. S. How go~ are the numbers for cigardtte tar at p~dicting deliveries of carbon monoxide, hydrogen cyanide, and acrolein. J. Toxicol. Environ. Health. in press (19~). Tt05280037