Tobacco Documents Online

Z~r-'.'E.~T~VE MEDtCthE IO. 301-,3tS flg~ll ~" ~TEVEN ~MAN.~ AN~ ~ATRICR ~,r£, A'v~, ~brA 1(ff117, ~Amerirtm Cum'ur .S'.t it'l). 777 Thh'~l A~ emw. New A sur~y ~ tobacco a,nd a]~ohol consumption was conducted among ~onins~ ~u.~ional~¢d men and women in 65 d~parlemc~s in France. Over four-fifih~ ~omen preferred bond tobacco products Younger smokers ofbmh scxe~ ~tcfcrrcd cigarettes, while ~/dcr responde~s smoked nonfihe~ed ones. Filter usage among men. but nol women~ increased with educalion.~ Inhalation was more prevalent among..men than - women, among the young ~han the old. and among the more ~ducalcd Ihan the "'Drooping~" or carry'~n~ u righted cigarct=es in ihe moulh withoul inhaling, wu~-a practice ~ore common_in olde~, less ¢ducaled mab ~mokcrs. These data also provide some evidence -- in support of the hypothesis ~hal black IobaccG. wi~h its higher pH. provide~ unprmonated _ nCco~ine wSi~ is abso~ed oral thus inhalation of the smoke deep less/mpo~unz ~o ~hc smokcr..The observed differences bc~ween.lbe smoking practices uflhe _ French and those in Ih¢ Untied Slates and Umled Kingdom along with the relaliv¢l)' higher alcohol consumplion holed ihroughoul France. may partially explain ~he lower rules of lung cancer and higher ra~es o~ larynx, esophagus, and oral cavhy cancer found in French men. INTRODUCTION For at least 20 years, lung cancer has been the most common neoplastic disease in many Western nations, in line with increases in per capita cigarette consump- tion. France, however, ~ains a relatively low lung cancer rate while rales of c¢~ain other ~obacco-related cancers, mos~ notably those of the moudL larynx, and esophagus, are unusually high among French men (Fig. 1) ( 10, 1 i). The lack of congruence in morality rates of these four cancers among the United Kingdom, the United States, and F~ance cannot be explained by differences [n tobacco consumption alone. We hypothesize tha~ the me~hod of cigarette smoking plus alcohol consumption conl~bule Io a pa~ial explanation of the internalional ratio- lion in rates as well as the differences in ~he ra~¢s ~[ween ~he sexes. The popularity ofcigare~les made wi~h black iobacco and lhe habit of holding a Ill cigareHe in the mouth without inhaling conslilu~e major deviations between the smoking habits of the French and ~hos¢ of the United Sta~es, United Kingdom, aad olhef nationalities. These and other di~¢rcnces were examined in a retro- spective study of the smoki~ ~d drinking habits of a s~ple of lh¢ French population. In~erest was focused ~ot only on the various £o~ms of ~ob~cco used but also on inhalation practices, which a~ ~h.eugh~ to be reJated black or blond tobacco smoke. An additional objective was Io invest~a~e T1~52800~2

~ 3~02~ ~x~DER E'F relationship between tobacco usage in France and various demographic variables such as sex. residence, and socioeconomic status. It was anticipated that such a survey would not only provide a good description of the French smoking pattern, but also aid in a rationalization of the varying rates of" tobacco-related cancers. Such data could help to make projections as to how changes in smoking habits might lead to changes in rates of these diseases in France. MATERIAL AND METHODS Data'were ctdlected by means of a French version of the standardized Com- prehensive Tobacco Queslionnair~ used by the American Health Foundation in the United States for the past l0 years (17). A team ofintervjewers experienced in meclical/scientlf'tc research, were specially trained in administering and coding the questionnaires. The questionnaires were checked for consistency, clarity, and completeness by one or'the collaborators (P, C.) who also served as consultant to the inter.viewers in the field. T105280003

TOa~A, CCO, ~SA~,~ |~ FRANCI~ 303 Between June and September 1976, more than 4,000 persons, between 20 and 80 years of age. were randomly selected and screened t~r study acceptance and potential interviewees had a condition~' which precluded participation in ~he study. Although no attempt was made ~o obtain a geographical probability sample, the demographic characteristics of the s~udy group were su~ciendy similar to those of the general French population to permit generalization (Table l). The major regions of France with the greatest population density, except strictly tourist areas, were included in the survey. That is, of the possible 95 administrative d~partements in F~nce, 65 ddpartements were represented in the study. The fol- 'lowing regions eea~bu~ed at I~ast 3~, of study cases: Marne (12.1%)Paris (12.~), Hauts de $'ei~e (11. i%), Vat (10.~,L Landes (9.2%), Essone (7.~), Gironde (3~,6%'), ~nd ~a~nche (3.~.). Thus, both rur~,l and urban area residents ~e q~e, saoae,~. ~e;¢au~-e ~4~ o~f.~e ad~lt ~en in Fm~ee smoke compared with 2~ of adult French women (3, 12), close to two males for each woman were incl'uded in the study (2~163 and 1.,2~, respectively). . RESULTS B~ckgromtd Data For each seE, a~proximately one-third 0fthe subjects ranged in age from 20 to 39, one-third from ~ to 54, and one-third above 55 years of age, All but 7% of the respondents were native born; 5~ were born in urban areas, and 62.8% spent their adult lives in uEban areas. Of the 246 immigrated to France, 85 ~35%) had done so before the age of 16, the average age at which smoking began. Thus, the great majority of smokers began smoking when living in F~nce, presumably with locally available tobacco products, and adopting local smoking habits. Approximately equal proportions of men had completed 0~6; 7-9, t 1-12. and 13+years of school. For women, ~ had 7- 12 years of schooling, with 2~ each above and below this level. Seventy-nine percent of the men (1,706) were employed at the time of the survey compared with 54% of the women (693): almost one-third of the women indicated housewi/~ as their occupation. Types of Tobacco Prod~cts Used Respondents were asked which specific combination of cigarettes, cigars; and pipes they used and whether they had ever used snuff or chewing tobacco. Ciga- rettes we~ "'ever" smoked by 82% of men and 56% of women, either alone or in combination with cigars and pipes. Sixty-six of the 2,163 men (3%) had smoked pipes exclusively, and 4.3~ smoked cigars or pipes exclusive of cigarettes. Ex- cept for 17 women who indicated use of cigars or pipes in addition to cigarettes, women either smoked cigarettes or did not smoke at all. As expected, pipes were smoked more o~en by older men (4.5~ at ages 55-89) than younger men ( 1.2~ at ages 20-39). Chewing tobacco had been used by ~ men I1.~) and 6nly one History of alcoholism or a psychiatric condition which might affect reliability of response. TI05280004

TABLE I I DI.~I PdllU I IoN uI- THI: DEklOI;tIAPHIC CHAItACTERISTI('~i OI-2 t'tlE ~'F,tlDY (~Kf)UP .a, ND ~RAN('E Study 8roup , ~ F~unc¢ ] Ivlalc (~'~) Fcmalu (~:~l "rol~l ¢,'; ) Male g:; ) Female (rH 20- 29 40 , 49 6O - 69 70- 79 Tolal Ufhan M~nilal b|alll~ $i))81c M:~l'ried Widowed I)i~orc~d/Scp~ alcd 18.8 18.0 13.9 14,5 23.0 21,9 19.1 19.5 14.5 14.7 10.7 11,4 2163 1290 Occtll~ltllOnal dit.s'-ificatitm Prt*l~.~mitmal 12.8 13.5 ,1| u,,i.n¢..is manageriul 8.8 4.3 "I'¢chni¢:d 14.2 9.2 Cl¢.rical/'iales~re I;tted 7.6 34.5 Skilledf-cmiskilled 46.5 29.7 lhv~killcd I0. I 8.7 Tul;.tl 1706 693 63 3453 . 16.0 ; ~70.0 'lO.O 4.0 3453 , i 24.7 212 18.7 16.1 19.7 18.0 13.7 13.6 13.5 14.9 9.7.. 16.2 16.9 millio,I 18.36 million 13.7 18.3 6.0 2.8 13.7 7.5 11.2 34,7 48.9 22.5 7.6 14.1 K1.62 million 8./3 million Soon'e, Relk. (3, 7, 8,1. 714

woman. Three-fourths of the men who had chewed tobacco were 60 years old or o,ye~. $;~ff use, either by nose o.r mouth, w~..ack~o~tedged by 326 ( The study defi.n,d nonsmokers as persons who had never regularly smoked any tobacco product; exsmokers as persons who had previously smoked cigarettes regularly, but had not smoked for at least one year prior to interview. Current cigarette smokers were divided into those who had been smoking for 9 years or less (termed recent smokers) and those who had smoked for at least tO years (long-term smokers); the remai.nd6r were persons who had never smoked ciga- rettes but who had ever smoked ciBars and/or pipes (Table 2). Sin-eking status v~ed markedly wifh sex, age, and. social status. Women were moire than three times as ,li,kel~ as. cae.~, to. be ttonsmokers (44 versus I4oA,), mad >am~rl$ the smo~e~;s~, ~.~¢~ ~et~e ~,~..~0~Les a~s l,~k~t:y to ibe current Iong-~rerm smok- ers (69' v~e,rsus 5J%),t~an women.. More than twice as many women than men were categorized as ~ecent smokers (20 versus 9c~). A slightly higher proportion of the female "'ever smokers" had given up the habit than the males in the study (25 versus 2,~). The proportion of men who had never smoked was between 10 and 20c/b in all age groups: t7%. bellow age 30, dropping to 12% at ages 60-69, and. increasing to 14% above age 70. Between ages 30 and 70, male smoking habits vary little with age,. except for a slight increase- in smoking cessation as the males grow older: As in other countries,- the smoking patterns differed between the men and women studied. That is, the proportion of female "'never smokers" increased steadily from 27% at ages 20-29 to 78e~ above age 60./he proportion of long-term smokers increased steadily between 20 and 40, slightly between 41 and 50, and decreased much more sharply than for males after 50 years of age. In all "~ge groups, among those classified as long-term smokers, proportionately fewer of the women than the men smoked nonfiltered cigarettes, TABLE 2 TOBACCO USAGE OF FRENCH MEN -~ND WOI~I~N ]h rkRVII'.WI-.D IlL I Wkk,'~ JIJ~l: ),h I) SI:PrI'.MIIER 1976 blales Femules N" "; .V %: Smoktng status Never smoked 301 13.9 570 44.2 Ever smokers Icigarettes only) [769 81.8 719 55 7 Cigartpioe smokers 93 4.3 t O. I Total 2163 100.0 1290 100.0 Cigarette smoking hi~tory Exsmokers Recent smokers Long-term smokers Total 386 21.8 182 25.3 165 9.3 tO4 20 0 12t8 68.9 393 54.7 1769 IO0.O 719 ll~kO TI0528000~

W Y~,D E,R lET ~n~ a~e~agea m:~r~ ~be~g~n smol~i~ng a@~p~,o~i~m~a~l~y ,2~ yea~s .earliest tha~n, women, and for both sexes a trend toward initiating the habit at younger ages is apparent. The.differenc~ between the sexes in average age-began-smoking diminishes to 1,5 in .other countries, smoking habits o~young women are becoming more similar to those of their male counterpa~s [Table 3). Smoking habits also varied with socioeconomic status. Among the men. skilled workers had the highest rate of cu~ent long-term smokers (72%), and professionals the lowest (48%). The reverse was t~e ~or exsmokers: 25% of professionals, 16% of skilled and semiskilled, and 8% of unskilled; similarly for pipe/cigar users. This situation was reversed for women, however. The largest proportions of nonsmoR~rs were found in t.he skilled (55%), semiskilled (4~), and unskilled (45%) groups a~d the lowest (27%) among professionals. Professional and ~u~,ne~wo.men. w,ere more likely ~ ~e long, term smokers than skilled or Quantity Smoked per Day .Men consistently smoked more cigarettes than women. While the mate- female differences remain about the same across ages, the absolut_e quantities smoked "drop by approximately 25% in the oldest group--a possible economic fac.t.0r., or . the fa~t~ that heavier, smokers..have died before.reaching old age. Below age 55, - .... men smoke approximately or~_e packeach day_,, and womer~ smoke three-fourths of ~ a pack; or 15 cigarettes per day (Table 3). These data are-also in accord with ~ ~ consumption patterns in the United States (17). ~ Clearly, cigarette smoking among the lower socioeconomic classes in France is TABLE 3 SELECTED CH,kR~,CTERISTICS OF CURRENT LONG-TERM FRENCH SMOKERS'~ BY AGE ~,HD SEX Age Age began Years smoked Years smoked Quantity smoked smoking nonfilters" filters" per day Average N Average N Average N Average N Males 20-39 16.4 287 14.7 191 11.7 160 .,21.1 287 40-54 17.8 511 26.0 427 15.4 261 20.8 511 55-89 18.8 420 41.5 d04 17.4 133 16,2 420 Females o 20-39 17.9 114 12.9 41) 13.1 98 15.7 114 40-54 22.0 203 18.?. 106 17.8 172 16.4 203 55-89 25.4 7~ 28.0 54 21.7 55 11.9 76 " Current smokers of cigarettes who have ~moked for at least I0 years. ~ Includes those who have smoked nonfiher for at least one year. • Includes those who have smoked filters for at least one year. TI05280007

~ncouraged in men ~nd discou~ged in women. In fact, until 1976. free cigarettes were rout~,nely disteibuted tO m~l'~s du, ring, Natio~l: Servie.. However. among the .t~e go~n~of l~rg~-scate ~ssadon. as in the ~n~d~ States (I~. Blond ~'ersus Black Tobacco Cigarette brands were coded according to a list prepared from the French gov- ernment's Liste Officielle ties Tabacs: differentiation into blond or black tobacco was made possible through supplementary marketing information. The sexes differed greatly in their preference for both filtered and nonfiltered ciga.re~tes and for black or b|ond tobacco. Men who listed a preferred brand chose nonffitered brands, over flittered (58 to 42~). while women preferred filters (77 to. 23%): me.n, o~'erwh¢lmingl~y preferred bl.ack tobacco, to blond (86 to 14.c-/c) while v~om'~n, sJ'ig:htly prefeere-dr blond (St to 49:%). The black tobacco brands most often smoked, whether by filter or nonfilter smokers, were Gauloises (the brand pro- vided during conscription), followed by Gitanes. These two brands accounted for 58% of the market share of all cigarettes among men~ and 48.6% for women. Smokers of black nonfiltered cigarettes tended to be older than either the black or blond filter users; of men who smoked nonfil~er Gauloises, 44% were aged ~.and above, while-,42% of nonfilter Gitnhes smokers were tiged 39 and below. Brand attd Education Males. There is an observable gradient with education in choice of tobacco product (Fig, 2), The percentage of persons [~sing blond filtered cigarettes in- creased from 5~4% among those with 6 years or less of education to 20% among those with at least 15 years of formal schooling; an obvious reflection that better- educated, more affluent groups can afford the more expensive imported Mond brands. Similar increases were evident with-black filtered cigarettes and the trend reverses for black and blond nonfilter cigarettes. Women. The trends among women are similar but less pronounced than those seen for men because the overwhelming majority t77~) of women smoke filtered cigarettes. Among these filter cigarette smokers, more than 60% smoke brands with blond tobacco. Nonetheless, as educational attainment increased, so did the proportion of women who smoked blond filtered cigarettes (39.3%, among women with 1"--6 years of schooling to 43.5,% for those with 15 or more years of educa-~ tionL Among the nonfiltered brands, 12% of women smokers specifically smoked Gauloises nonfilters, compared with 39. 1% of male smokers. As years of school- ing increased, a steady decrease was seen in the proportion of women who smoked nonfiltered cigarettes, in general. Occupation and Filter Use Choice of filter and tobacco type also varied considerably by occupational status. Among_male smokers, nonfiltered Gauloises were preferred by 54% of unskilled workers who smoked, but only by 29% of male professionals and execu- Ti05280008

308 ",.','Y N D E R ET 80. ¢------o 6L~,CR FILTER o..-----o BL~.CK NON-FILTEg ~E~,R$ OF EDUCAr¢ON Fro. 2. Tobacco type and fihcr preferences amon..q French mal¢s ~y e~ucafion. fivesJ OF {his Imter ~roup. 2~C; smoked the more expensive blond brands, compared w~th 6~ of the unskilled workers who did so, Women who were either professional or bfisincss executives, or whose hus- bands were,, preferred blond filter cigarettes to Gauloises -nonfilters (54-versus 12% 9f smokers), while women in. the unskilled catego~ showed the same pFefer- ence:but to-a lesse~-degree (38 versus 15~). Cm'u Poper (Papier Mais) Smokers were asked whether they ever used papier reals, a slow-burning wrap- per. Only 6.3% of male smokers, and 0.7~ of female smokers had ever used it. Usage was noticed mostly in smokers above age 40. In accord with reported inhalation patterns in the United States and United Kingdom. men on the whole inhaled more deeply than women and twice as many women than men (31 versus 16%, respectively) claimed not to inhale. A note- worthy difference in ages of non~nhalers b~tween men and women is-demon- strated in Fig. 3. As in other countries, the proportion of noninhalers increased with age for both men and women, and the reverse was true for deep inhalers. The amount of inhalation reported by French smokers in this study is compared in Table 4 with that reported in British and American studies (9. 18). While responses to qt,estions on inhalation must be viewed guardedly because of their svb.iective nature. Ihe questions in the U.S. stndy were identical to those used in the French study and. therefore, the results should be comparable. The majority of men who smoked nonfilter blackcigarettes inhaled less than male nonfilter smokers in Britain or America. A possible explanation for this is that the smoke is T105280009

]rO, BACC,O U:SAG~E IN FRANCE 30,9 ~0 Z0 I0 0 DEEP IION INHALERS INHALERS ~ HALES o--------o FEH~,L -ES o-. FIo~ ~ .,I,~h~.~tion practices of French men and women, by age, 1976. less acidic than either blended American cigarettes or Engligh variety cigarettes. Under conditions of high pH~ a significant fraction of the nlcot~ne is unprotonated, and therefore, re.ore easily absorbed through the oral mucosa (I, 2)..Thus, a French smoker of blacknonfiltered cigarettes may~not need to inhal~ as deeply in o~der to achieve a "satisfying'~ blood level of nicotine. _ " " The difference in inhalation patterns between the blond and black filtered brands is a reflection of the younger smoking habits, i.e., proportionately more young smokers smoke filter cigarettes and proportionately more young smokers inhale the smoke. For both sexes, proportionately more smokers of black filtered cigarettes inhaled the smoke deeply into the chest. Drooping In contrast to Americans, many French smokers hold the lit cigarette in the mouth without inhaling the mainstream smoke. It is known that sidestream smoke (that combustion product diffusing into the air directly from the main fire without passing through the cigarette) can contain more toxic components than the mainstream 12), The frequency of the "'drooping" habit was asked and the an- swers differed markedly between men and women. Fewer than half or the men (42.2%) and three-quarters of the women (75.3~) never held the cigarette in their mouths without puffing on it. One-third of the male smokers responded than such as the case "halF" to always, Compared with 1 I..~ of the women. The frequency of this "drooping" habit increased with age among males with no such trend appar- ent among the women (Fig. 4), Ofmajor eoid.emio!ogie interest is the fact .that not only do inhalation and "'.drooping" practices-vary with age. but they also vary in opposite directions. That is. "'droopi.rtg'" is a habit of the older mal'e smokers tas in the use of corn paper, which greatly facilitates this habit), and this habit diminishes the amount of T!05280010

CO,MPARI.%ON oF [NIIAI.A'i'ION PRACTICES UY SEX AND TYPE OF Males Filzered (HI Nunlillered t,~'~ ) Fihcred (,~) ,Inh..d¢ ~ Iut 46 inhale u l~ir amount 32 lphalc - Ihtle 17 Do nuz inhale 5 55,, 22 23 32 14 31 8 15 United $tate~' Males Filtered 4"~ ) Nonfillered (¢:; I Filtered .Maids 53 6 7 7 58 39 24 " 33 6 il 6 6 6 12 France (t, his sludy) Females Females Fil|ered Nonfillercd Filleted Blond (91.) Black (~) Black (¢;.) Blond ((~) Black (c~) Inhale de~ply 33 $6 Inhale parlly Io ehesl 23 19 Inhale back to Ihroal 15 7 Inhale, deplh unknown [9 6 Do nut inhale " 10 12 Total numbers 181 357 26 15 37 792 Females 23 49 2O 17 17 14 6 6 34 13 272 143 Ni ~0$1tered ~ ~, (,:~) " Source. ReI: (9). * Source. Ref. (18).

mainstream inhalation. Because~.women have relatively recently become major users of tobacco products, however, t~ieir smoking h~ibits are more s~milar t0 those of the women in. the United States and their younger male counterparts. Alcohol Consttmption The use of alc,ohol in conjunction with tobacco is known to increase the risk of "cancers of the mouth, larynx, and esophagus (18). These survey data confirm the stereotype that wine is the most frequently consumed alcoholic beverage in France ¢Table 5) among men and nonsmokers. Some additional, nontabulated data TABLE BI~V|;RAGE CONSU).II*I ION Alcohol Males Females Nonsmoker Long-term ~onsmoker " Long-term .V ¢~. ,',' C~ ;V e4 .V ~'~ Wine only 123 41 409 " 34 207 36 101 26 Wine and beer 22 7 91. 7 8 I 17 4 Wine anct licluor 47 16 .~17 26 68 12 82 21 Liquor or beer 18 6 71 O 56 10 52 14 All three 20 7 191 16 1,1: 2 2l 5 Never drink 71 24 140 I I 220 39 120 31 Total -30 t 12t9 570 393 "'>~ lOyenrs. I T105280012

3,12 "~VYNDER ET A~.. are of i,~tcrest: the highest Bro,port~on~ of ndr~dr~n,kers was found among recent " This group is almost entirely below 30 years of age. A~er recent smokers, nonsmokers had the largest proportion of nondrinkers (24%). But 4[% of ~,~s ~ank w~ne (alone). and ~ t~ of t~:m d~ wi~¢ w~th~r a~to~ or in: ~0¢~P~i~t~o:ns ~t~ atcontrol' eon,~u~-ption /'~.y a|so account For variation ot" alcohol-related cancers within. France. Table 6 contrasts the alcohol consumption patterns between Manche.. a largely rural area in Brittany, and Paris. and shows a marked increase, in the former, of not only liquor consumption, but also usage of all three types of alcoholic beverages. Manche and neighboring Calvados have very high rates of esophagus cancer, and are also regions of much higher con- sumption of alcoholfc cider and certain other beverages than Paris (13, 14). Epg4tentiolt) gic Considerations The .p~,~,n.t study has con_cot:ted itself with the epidemiology of some smoking- .v~:t.~d,; vea~ir]~l~l~:i.~ Frane.e. Whitle th~ pere~n.tage of nonsmokers d~.es not differ muc~h~ ~.a~~ng ~he United' States, United Kingdo,m, and France, there are consider- a.bl!e differences in inhalation, "drooping,." and drinking practices among older men and women in France (3, 7. 13. 14k It is possible that the lesser inhalation practices in older French men are the result of" the high pH of French_ black tobacco, which y!eld a larger amount of free nicotine, and thus contribute to a lesser likelihood of deep ~nha!adon, .A" !so, the habit of"drooping." which is .rein--. tively common among older men in France. makes deep.inhalati0n less likely, although possibly increasing passive inhalation of sidestre~.m smoke, We suggest that the lesser degree of inhalation and the greater proportion of"" drooping" and alcohol drink.ing among older Frenchmen are factors associated with low rates of lung cancer and high rates of esophagus, larynx, and oral c~avity cancer.. Prolonged exposure of oral cavity and esophageal tissue caused by the retention of the cigarette in the mouth with the increased exposure to sidestream smoke could account for the international differences in these mortality rates, The Manche Paris Wine only Wine and beer Wine nnd liquor • Liquor :rod beer truly All ~hree' Nondrinker Toml I I 17.7 73 28.6" 2 3.2 22 8.6 5 8. I 39 15.3 2 3.2 23 9.0 31 50 0 .'15 13.7 I I 17.7 63 24.7 62 255 ~ ln¢lude-s ak-ahol[¢ eider. Tt052800t3

fivefold difference in laryngeal cancer rates between France and the United, States poss~bth,laes. Not o'ul~y do wome~ hav~-~s~ tobacco and, alcohol exposure but deep inhaling and "'drooping" is encountered far less frequently among women, it is suggested that these practices could account for the noted sex dift~,rences in tobacco-related diseases. The higher nitrate content of black tobacco may also affect the reported lower lung cancer rates in France since experiments indicate that tars obtained from higher nitrate tobacco cigarettes have a reduced tumo¢ yield in mouse skin (61. It is signifiea~.,t tha:l; :,~i~'~ lg~ls of A~merican~ c gare,~tes have been rising in, .recent years, during t~ period, or' decline in mr and nicotine. It also appears that these ca,g~r rates ha~ ,~~,~ :b:_~n [nerea~s~d"d~etectably b7,~ the:, h,ighe~" n trosamine content animal~, some ~i,t'~s~,ami~nes ha~e l~e'en found tt~ be organ specific for the e$0phagus--,a finding which might also apply to man. and thus contribute, to creased rates of esophageal cancer. It is reasonable to asst~me that tobacco and/or tobacco smoke are the sources of the initiating carcinogenic stimuli, and that alcohol (ethanol) facilitates the reactivity of some tobacco-assoc~a!ed initiator. ~But'clearly such a relationship does not obtain tbr the lung and, therefore, cannot, be associated with the_low rates of lungcancer in France, These data co~afirm the results ofoth~r studies~vhich imbilcate th~ highe~ iotal alcohol intake in France, in the. development of Cancer of the upper alimentary tract, particularly in llne with the fact that alcohol has been shown to significantly enhance the development of these cancers in smokers 1 18). Furthermore. we may conjecture that if tobacco smoke is inhaled less, more smoke is retained within the upper alimentary tract and may thus contribute to the reported higher rate of upper alimentary tract cancers in France. In Franc'e, as in other countries, both smoking habits and the cigarettes them- selves have changed over the years. Table 7 shows changes in the tar and nicotine le.vels of French, British, and American cigarettes, both filter and nonfilter, over a 15-year period t9, 10, 141. The decrease in tar, even in nonfitter 6igarettes. as well as the increased preference in all three countries for tilter cigarettes is likely to have an ameliorating effect on lung cancer rates. A lower lung cancer risk among smokers of low-tar cigarettes has been noted in the United States by Hammond as well as our group t 161. On the other hand, reduced tar and nicotine levels might lead to an increase in deep inhalation of the smoke, and thus counteract some of this reduction.in rates of tobacco-related diseases. Finally, the rapid rise in the. proportion of French women who smoke may cause a rise in their lung cancer rates, as seen in the United States and England t4. I0) C,ONOLUSION The lower rate of lung cancer in France, as compared with England and the United States. is a~sociated with the lesser extent of inhalation among older French smokers and the greater prevalence of "'d?ooping'" the cigarettes, The T105280014

314 WYNDfiR ET AL. TABLE 7 BRITISH. AND AMERICAN CIGARETTES. Nicotine 2.0 2.2 1.4 1.9 France 1976 14 24 1.0 1.5 England 1975 17 25 1.3 2.0 United States 1976 17 26 1.1 1.6 S¢mr¢'e. Refs. (9, IO. 141. higher rate ot~ oral cancer in France compared with these two countries is consis- ten~ with+ ,the 'high cons9mption of ateohol in, Fra, t~e+e, and, the fact that propor- ,ti++ +~t¢l~y nao~r,e older Preta~:l~rncn held the lit c~g,a++e~ttre in +he mouth for extended periods of time. Yo,unger French smokers inhale more and "droop'" less than older smokers,. perhaps as a consequence of lower tar and nicotine levels in newer French ciga- rettes. It remains to be determined whether the future rates of tobacco-related cancers in France +will become more like those of the United States and United Kingdom as the French adopt re.ore comparable.smoking habits. That is, it seems possible to predic_t that not only will the sex differences in rates diminish but also. that the mortaliW rates for lung cancer will increase (+while those of larynx+( esophagus, and oral cavity might decrease). Continued monitoring of tobacco- related diseases in France in relation to cigarette consumption is in order. ACKNOWLEDGMENTS We wish to thank the following persons for thei~valuabl¢ assistance in administering the queslion- naires: Michel Froidevaux. Victor Marengo, Phillippe Dard. Phi.IIippe Girault. Jean Pierre Huc. Herve Juille, Jean Lepagno, Pierre Lerminet. Marie Louise Lorillot. Stephan Thorn. Gilles Vuillierme. REFERENCES i. Armi'tage. A. K.. and Turner. K. M. Absorption of nicoti~ae in cigarette and cigar smoke through the oral mucosa. N, ture (Lomlonl ~26. 1231-2332 (19701. 2. Brunnemann, K. D.. and Hoffmann. D. The pH of tobacco smoke. Food Cosmeud. Toxico/. 12, 115-21 f 1974). 3. Department of International Economic and Social Affairs. Statistical Office. "'Demographic Year- book," 29th issue ( 19771. United Nations. New York. 1978. 4. Doll. R. Surveillance and monitoring. Int..I, Epith,mlol. 3, 305-314 t 197,1'L 5. Hammond. E. C.. Oarfinkel+ L.. Seidman. H.. and Lew. E. A. Tar and nicotine content ofeiga- retie smoke in relation to tlealh rates. Em'/e+m. Re:+. 12. 263-27,1 (19761. 6. Hoffmnnn. D. Unpublished da~.~. 7. International Labor Office. "" 1979 Yearbook of Labour Statistics:' 29th issue+ United Nations. Gene,~a. L980. ~, Ledez. P. Personal ¢ommuoieation. 9. L¢¢. P. N. (Ed.) "'Statistics of Smoking in the United Kingdom. Research Paper I." 7th Tobacco Research Council. London. 1976. Tt05280015

10. Levin, D. L., Devesa, S. S,, Godwin, J. D., and Sil~crman. D. Cancer Rale~ and R~sks£" 2nd ed. U.S. DHEW ~bl. (NIH) 7~6~1. 1976. TI05280016