Tobacco Documents Online

• l t ri L. _.=z;... Some~~ Recent Fi~nd6~ngs, Concerning, Cigarette Srnoking E C. Hammond, L Garfinkel, Hh Seidman and E. A. Lew Department of Epidemiology and Statistical Research American CancerSociety, New York. New York 1i00'1'7' © An enormous amount of research on the effects of smoking has been carried out by.' many izadepend'ent investigators in many different countries. The epidemfioiogic findings have been so similar in' various studies-especially prospective studies-that data from any one of' them can be used to illustrate the major results. The literature has been reviewed so often and so well that we will not repeat the process' here. However, we will start by dis- cussing a few salient points. First, and most important, death~ rates are higher in smokers than in non- smokers sutokers andl they increase with degree of exposure to tobacco smoke. Among the diseases involved in this relationship are (1) lung cancer and cancer of several other sites, including the lip, tongue, mouth, larynx, pharynac, esophagus, and urinary bladder; (2) coronary heart disease, stroke, and aortic aneurysm; (3) chronic bronchitis and emphysema; and (4) several other diseases, including peptic ulcers. The latter was most elegantly investigated in a clinical trial carried' out by Sir Richard Doll and his associates (Doll et a1. 1''958) . Figure 1 is based'upon data from one of'the prospective studies and shows' lung cancer mortality ratios in relation to types of smoking. Note that althoughh death rates from this disease are considerably higher among pipe and cigar smokers than among men who never smoked' regularly, they are far higher, among cigarette smokers. This appears to be due to the fact that the majority of pipe and cigar smokers do not inhale the smoke or inhale it only slightly, whereas the great majority of cigarette smokers (especially heavy cigarette smokers) inhale the smoke to a moderate or deep degree. Among those few pipe and cigar smokers who consciously inhale the smoke, litng cancer death rates (as well' as death rates from coronary heart disease) are as high as the death rates for cigarette smokers. On, the other hand, death rates from cancer of the lips, tongue, mouth, and esophagus are as high or higher among pipe and' cigar smokers as among cigarette smokers, regardless of degree of inhalationof'tlle smoke.. 73-1Z50. Hammond.' E C.; Garfinkel.,Li.: Seidman. H.; Lew. E A. Sotrte Recent Findings Caonoertting C7gtrette SawkirtG In: Hiatri H. H.; Watson. J. D.; Winsten,!! A- (Editors). Origjnr of Hwtmr Canoe. Book'A. Jneidenct of Canrer in Huni Cold Spring Harbor Conferences on, Cell Proliferation, Volume 4. New York, Cotd Spring Harbor [iboratory. 1977, pp. I Ul«111II: Yoticr. This matenal may bepr<otected byoopyright laov. (Tide 17 US code) 1r 1D1 I I. , IJ l It t r ~ © * m M I Mi ~ m © t t<~~ ~ a ~ r. :r. .Vg~w .,.5- ['a t ~'«`? . ~, Q . 17 ,,..... J~ 'w ~ ~':~ ® 4 ? Y

1 l:24 lioe r,.-.:-... - N.S1R P1PE'.. qGeR : pGARETTE GGGRET7FONLY ONLY. ANDoNL1f t1rHER' Figure 1 Lung cancer mortality ratios in relation to types of smoking. Lifetime smok- ing,history' (including ex-smokers). Figure 2' shows mortality ratios ( al! causes of' death, combined) in relation to both number of' cigarettes smoked per day artdl the age at which cigarette smokers began to smoke. Death rates increase with the number of cigarettes smoked per day, and, among men who smoke the same number per day, death rates are considerably, higher in those who begani under the age of' 15'than in those who began after the age of 25. - These two variables and'' degree of inhalation are all interrelated. Men who began to smoke at an early age tend, in later life, to smoke more cigarettes per day and! inhale the smoke more deeply than those who begani at an older age. Note that on, this graph no point is shown for the mortality ratio, of' menn who smoked only 1-9 cigarettes per day and began to smoke under the age of 1'5.11ias is because there were so few such men-and, conseqtzently, so few deaths in this category-that their death rates could not be estimated reliably. The remainder of this report is based upon new an& previously unreported data. 0 1-9 1 qHi 20•39. CfOJ1RETTES PER DAY Figure 2 40s. Mortality ratios in relistion to number of cigarettes smoked per day and, age smoking began. Total deaths are shownifor men of age 55-64. _rntiee: This material may be Proteoted byoop,,vrigNt law. fTWe tT'tDScode ..--~ --~•--

. Smoking 103: Dose-Response At the present time there is consid'erable interest in the shape of dose-response curves for human beings exposed to various noxious agents. The principal question is whether in attempting to estilmate the effects of very low dosage from the known efl;'ects of relatively high dosage it is valid to assume that the dose-response curve is Iinear. Figure 3 shows age-standardized death rates for coronary heart disease in relation to l the number of cigarettes stnoked'i per day, for each of three attained age groups. The points certainly do not fall along al straight line for any of the three age a oups. For men in attained' a;e group 70-79, the coronary heart disease death rate was lower for those who srrtoked 40' or more cigarettes a day than for those who smoked 21 to 39 cigarettes a day. This may have been due to selective mortality. These figures are based upon the mortality experience between July 11, 1966: and June 30, 1972 among a large number of, rnen who were enrolIed' in 1959-1960 in a prospective study by volunteer workers of the American Cancer Society. All of the subjects included here answered a follow-up ques- tionnaire between October 1, 1'96S' and June 1966. Alli o[l the smokers were currently smoking cigarettes at' the time they answered the repeat question+- 4il 0. hf ~ *! 20 ~. 21-M p~. u. (N.6)~ 0131 (2l:2)i Ulill CIGARETTES PEFt' DAY I FigWre 3 Age-standardized death rates for coronary heart disease. #Maie. TSir matenial may be protected brcopyright law. (Title ! 7 US code) .

104 E. C. Hammond et al. e a (NJ1 I12J (29.21 0"f CICAtiETTES PER DAY F+4 4N 20 tl~-39' 40 sls ~~ .. Figure 41 Age-standardized death rates for lung cancer. naire, and' none of them had ever smoked pipes or cigars regularly up to the time they answered the first questionnaire. They were classified according to the higher of the following two numbers: the number of' cigarettes they were currently smoking per day aG the time they answered the first questionnaire and the number they were smoking per, day at the time they answered the 1965-1966 questionnaire. Figure 4 shows age-standardized' hnng cancer death rates for the same sub- jects. The shapes of the curves (if they can be so~called') are quite different from the shapes of the curves shown in Figure 3 for coronary heart' disease.. It should be noted that the death rates indicated by some points on' this chart are very unstable statistieally due to smalPnumbers.. As described previously, the effective degree of'exposure to cigarette smoke depends upon the age at which a person begins to'smoke as wel!1' as upon the number of cigarettes he smokes per day. We are undertaking a further analysis taking both of these exposure variables into consideration. Tar and Nicot'ine. Some years ago a small committee of experts on the subject (U. S. Congress 1967) came to the following, conclusion: "The preponderance of scientific evidence strongly suggests that the lower the 'tar' and nicotine content' of ciga- -.rotier. T1his material may be protected bY,copyright law. (Titte 17 US code) w~y,

Smaking i 105 rette smoke the less harmful would be the effectsA (A short time later this was reported by the then Surgeoni General of the U. S. Public Health Service.), Their reasoning was roughly as follows: (1) Death rates from lung cancer, cancer of'severaliother sites, coronary heart disease, and severaI, other diseases increase with degree of exposure to cigarette smoke. (2) Many experimental studies have shown that material condensed fromi cigarette smoke (usual0y, called "tar") is carcinogenic when, applied to anitnals: (3) The known acute effects of nicotine upon the heart' and circulatory system suggest that the nicotine content of cigarette smoke is partly, if not entirely, responsible for the fact that age-speci'fic death rates are higher among cigarette smokers than among,nonsmokers. (4) Therefore, it seems reasonable to suppose that if'the tar and nicotine content of cigarette smoke were reduced, then the harm, done per cigarettesmoked wouldibe correspondingly reduced. ' The term "strongly suggests" included in the statement by the committee cited above iniplied'that the conclusion might be incorrect. The major counter- speculations were: I. If'the tar and nicotine (T/N) content of the smoke were reduced, most smokers might smoke more cigarettes per' day and thereby cancel' the benefit. (This speculation turned' out' to be incorrect when it was found that smokers who switch from high TIN to low T/N cigarettes do not usually increase the number of cigarettes smoked per day. ) 2. Smokers of'lpw T/N cibarettes might (consciously or unconsciously)i in- hale the smoke more deeply than smokers of high TIN cigarettes. If so, then their effective exposure to tar and nicotine might, not be reduced and their exposure to the gases in cigarette smoke woul&be increased. 3. It could be that gases contained in cigarette smoke are as harmful, if not more harmful, than the tar and nicotine content of the smoke. Furt;hermore,, it could be that, under certain circumstances, reduction in T/N is accom- panied by an increase in certain gases, most notably carbon monoxide. An increase in carbon monoxide might increase the risk of coronary heart disease. Therefore; if all' this be true, the net' effect of reduction in tar and nicotine might be an increase in age-specific death rates. c~c . .. Since that time, published evidence from two retrospective epidemiologic studies (one by Bross and' Gibson [1968] and' the other by Wynder et al.. [!1970]) has indicated' that people who smoke filter-tip cigarettes have lower lung cancer death rates than people who smoke nonfilter cigarettes. Concernn that this desirable effect migttt be accompanied by an increased risk of some other disease (especially coronary heart disease ) led us to carry out an iib- vesti;ation which we will'now report. The prospective study referred to, previously consisted of over 1,000,000 men and women in 25 states who were enrolled by volunteer workers of'the American Cancer Society between October 1, 1959 and March, 31, 1960 and' up~on~, enrollment~ answered a detailed', questionnaire. Appromiinately~ 9'8'.4~%% of them were ttacedi througk September 30; 1'965; and of those still alive at that time, 94'.9'% answered a relarcivelyshort follow-up, questionnaire. All'but a few of'these had also answered one or both of two previous repeat ques- tionnaires. Tracing,was then discontinued for 6 years. On October i', 1'971, we began what amounted to a new study. The subjects were 897,825 men and women in 23' states who had' been subjects of the first Notien This material may De pnotected ItiycoP,yrnght law, Mtie l7'US codb) I

P study and who~ were still alive as of September 30„ 1965. Of these; 98,4%~ were traced through September 30, 1971, and 92.8% were traced through. September 30, 1972. This report is confined to experience during two 6-year periods: Period 1, July 1, 1960 1 through June 30, 1966; Period 2, July 1, 1966 throughiJune 30, 1972. For both periods, it is confined to people who never smoked! regularly and people who were smoking cigarettes daily at the time of initial enrollment and had never smoked pipes or cigars regularly. For the second period, the report is further confined to people who answered the 1965-1966 question- naire and who were either still' nonsmokers or were stilli smoking cigarettes daily. In each questionnaire, cigarette smokers were asked, among other things,, to name the brand of cigarettes they usually smoked. Since the tar and nicot2nee content' of the mainstream smoke of'various brands of, cigarettes has been re- ported' from time to titne, we were able to divide the subjects into three sets, which we refer to as high TIN, medium TIN, and low T/N smokers.. This was relatively easy for the first period, since even though some m.anu- facturers marketed two or more types of cigarettes under, the same brand name, they could be distinguished by the presence or absence of a filter or by menthol. For the purposes of this report, we defined "high" T/N as 2.01 to 2.7 mg of nicotine and 25.8 to 35.7 mg of tar. "Low"' T/N was defined' as less than 1.2 mg of nicotime, and, with very few exceptions, cigarettes which met this qualification also delivered less than 17.6 mg of tar. "Medium" T/NN was simply defined as intermediate between high and low. Insofar as possible„we used these same definitions for later years. However, during this later time period, some manufacturers marketed under the same brand' name two or more types of cigarettes, which differed in tar and' nicotine content. For this reason, it is likely that some of the sub9ects who were placed in the high TIN category probably belbniged' in the medium T/N' category, and vice versa. There was far less difficulty of' this sort' in distinguishing the low TIN smokers from the other two groups. For the period 1966-1972, the three sets were distinguished as follows: (T)i "Higb"' was defined as subjects in the high category in the 1959-1960 questionnaire and as either high or medium in, the 1965-1966 questionnaire. (2), "Low" was defined as low in the 1!959-1960 questionnaire and as either low or medium in, the 1965'-11966' questionnaire or as low in both the 1961- 1962 and 1965-1966 questionnaires. (3) "Medium" was defined as all other smokers. Matched-group analysis was utilized. This process is similar to age standardization of'death rates except that the standardization is based'on age plus a number of other factors. Male cigarette smokers were divided into~groups, each group consisting of men who, at the start of a periodi were alike with respect tio age, race, number of cigarettes smoked per day, age they began to smoke cigarettes, place of residence (urban or rural), occupational exposure to dust, fumes, chemicals, etc., education, prior history of lung,cancer, and prior history, of heart disease. Female smokers were divided, into groups on the basis.of' all~ the above-named factors ~except occupationaexpbsunes: A group, so defined, was discarded if it did not contain at least one low 3'Ilotkr. This macerial may be protected by oopyright la w.,(Title 17'US code) ~

.~ w .. . ~~. s c±7± J- Smoking 107' High hledirein Low Sex Period TIN TIN TIN Table 1' Total Number of Subjects at Start of Period'. Male 1960-1966 63,063 ' 54,999 115:3601 Male 1966-1972 29,157 40.090 6832 Female 1960-11966 44,137 59,750 32,703' Female 1966-1i972' 22,909 49,193' 16,80.3' a ~ , ~. ~ ~ " ."l.r ~ TMM - ~I'h 4 4 ~ =.E . ~ . T/N, one medium T/N, and one high T/N smoker. Most groups containe& fewer subjects of one type than another type. For example, a group might consist' of 50 low TIN, 200 medium T/N, and 150 high T/N smokers, the ratios of these numbers being, 1, 4, and' 3, respectively. The ad jtated number of, subjects was in this case 50, this being the number of subjects in the type with the fewest subjects. The adjusted numbers of deaths were (in this case) computed by dividing the number of d'eaths in the low T/N subjects by 1, dividing the number of deaths in medium T/N subjects by 4, and dividing the number of deaths in high T/N subjects by 3. After similar caleulations were carried out for each group, the adjusted numbers were summarized over all groups. The logic of this procedure is similar to the logic of an experimentalist who first makes sure that the animals in his experimental sets and the animals in his control sets are as alike as possible and then makes an adjustment for any difference in the number of animals in various sets.. RESULTS Table 1 shows the unadjusted number of'subjects at the start of each of the two time periods by sex and by tar and nicotine. There were fewer low T/N smokers than medium T/N and high T/N smokers. Table 2 shows the ad- justed number of'subjects, which by definition was the same for smokers of' high, medium, and low T/N'cigarettes. -The adjusted number of deaths is shown in Table 3. For both men and women, in both time periods, the adjusted number of deaths was lowest in the low TIN category and' highest in the high TIN category. Figure 5 shows thee same data in terms of' mortality ratios. The adjusted number of deaths in low Table 2' Adjusted, Number of Subjects . ~ Sex Period Number Male 1960-1966 1'4,68'8 Male 1'966-1972 6475 Female 1960-1966 30„L'76 Female 1966-1972 15,342 r .Yntiee: Ttiis material may beprotaeted bineopyright law.,(Ttle 17'NS code) » . ~ - _ . . .r t r . : ~ ~... .: r: .`. Y .... , 0 ®

Table 3 Adjusted' Number of' Deaths (Total' Deaths ). , Ser Male Male Female Female Total, Period' I3isJt TIN Medium T/N Low TIN 1960-1966' 1543.0 1394.4 1351.7 1966-1972 935'.2' 9'13'.7' 759.4 1960-1966 12534 1'117.1 10519 1966-1972 .1i00.3'.7 874.7 826:2' total 4'735.5' 4299.9 3991.2' TIN smokers ranged from 81 ~'o to 88 % of the adjusted number of deaths inn high, TIN smokers. Table 4 shows the the adjusted' nunzber of deaths from coronary heartt disease (C.H.D.). The numbers are reasonably large in each category. Figure 6 shows the same data in terms of' mortality ratios. In i both periods of time and in both men and womenj the adjusted number of C.H.D. deaths for low T/I*T'smokers was lower than the adjusted number of C:H'.D, deaths for high T/N smokers. . Table 5 shows the adjusted number of deaths from lung cancer. In both periods of'tizne and in both sexes, the adjusted number of deaths was highest for high T/N smokers and lowest for low T/N'smokers. Figure 7 shows thee same data im terms of mortality ratios. The adjusted number of lung cancer deaths of Iow TIN smokers ranged from 57'% to 83'% of the adjusted number of deaths of'the high T/N smokers. Obviously the amount of'tar and nicotine taken into the body per day de- Tl1'R;/NIC©T1NE'IN CIGARETTE SMOKE Figure 5 Total deaths in terms of' mortality ratios for high, medium, and low T/N cigarette smokers. .Notkr. This nnaterial may be protected br cooyrf ght law. (Ttle I7'(7S code)

' _ Table 4 Adjusted Number of' Deaths from Coronary Heart Disease Smoking 109 fJLi;lt Afedurnt Low = Sex Period TIN TIN TIN Ma1e 1'960-1966 69,6.5 ' 632.5 645.6 Male 1966-1972' 336.0' 345.6 274.2 Female 1960-1966' 315:7 277.5 257.41 Fem ' a1t 1966-1971 y 263 6 228 0 2'1S 5 Total totai, 1616.8 1483.3' 1392.7 pends upon the number of cigarettes smoked per day as well as upon the tar and nicotine content of each cigarette. To obtain some information on the relative importance of t'hese' two exposure variables, we carried out a second matched-group analysis (Table 6). This analysis was confined to two sets of subjects: (I ) subjects who smoked I to 19 high T%i~1' cigarettes a day and (2) . subjects who smoked 20 to 3'9 low T/N cioarettes a day. As shown in Table 6, the adjusted number of lung, cancer deaths was greater in subjects who smoked 20''to 39 low T/N cigarettes a day than in subjects who smoked 1 to 19 high T/N cigarettes a day. This was true also for total deaths and for C.H.D! deaths, but to a far lesser degree than for lung cancer as shown here. . Finally, in still another matched-group analysis, we compared low T/N smokers with subjects who had never smoked regularly. As shown in Figure 8; death, rates (ftom, all causes of death combined) were considerably higher in subjects who smoked!low T/N cigarettes than in subjects who never smoked' regularly. Figµre 9 shows~ that the lung cancer death rate was far higher in. low T/N' smokers than in subjects who never smoked! reguIarly.. MEN PERIOD I PERIOt1 2 I n. .W M L N M L N' M L H M L TA'RJNICOTINE IN CIGARETTE SMOKE Figure 6' Coronary heart disease mortality ratios. ~oticr. This matetiat tnay be'protected IbycoPynigbt law. (Title l7'US code) w. p

Table 5 Adp"usted'Number of Deaths from Lung Cancer Ser Period' 11F,p11 TIN Mcdirrnt TIN Low TIN Male 1960,-1966' 122.4'. 117.4 1101.01 Ma1tr 1966-1972 89.6. 84.5' 70.6 Female 196iD-1966 43.3' 41'.4'. 27.4 Female 1966-1972' 58.1 42.2 3i6:2 Total total 318.4 285.5. 235.2 1b TAR/N160T1NE IN! CIGARETTE SMOKE' Figure 7 Lung cancer mortality ratios for high, medium, and low T/N cigarette staokers.. Table 6' Adjusted Number of'Deaths from Lung Cancer 1-19 20-39 N Srs Period High Low TIN TJN cigarettes cigarettes ~ Male 1960-1966 363 63.9 ~ Mile 1966-1'972 24:7 33!8' ~. Female 1960-1'966 103 118.0i N Female 196i6-1!972 4 5 1313 ~ Total total . 75.3 129.5 UD Q 110, •Plaiee: This materrtal may taeprotectedl"pyrighclaw: (Title l7 US'code) «