[CANCER RESEARCH 50, 6502-6507. October 15, 1990]

Smoking and Drinking in Relation to Cancers of the Oral Cavity, Pharynx, Larynx,and Esophagus in Northern Italy1

Silvia Franceschi,2 Renato Talamini, Salvatore Barra, Anna E. Barón,Eva Negri, Ettore Bidoli, Diego Serraino, and

Carlo La VecchiaEpidemiology Unit, Aviano Cancer Center, Via Pedemontana Occ., 33081 Ariano <PN),Italy [S. F., R. T., S. B., A. E. B., E. B., D. S.J; Hormones, Sexual Factors andCancer Group—European Organization for Cooperation in Cancer Prevention Studies, Brussels, Belgium fS. F.J; Department of Preventive Medicine and Biometrics,University of Colorado, Health Science Center, Denver, Colorado 80262 [A. E. B.]; "Mario Negri" Institute for Pharmacological Research, Via Eritrea, 62, 20157 Milan,

Italy ¡E.N., C. L. V.¡;and Institute of Social and Preventive Medicine, University of Lausanne, 1005 Lausanne, Switzerland [C. L. V.]

ABSTRACT

A hospital-based case-control study of upper aerodigestive tract tumorswas conducted between June 1986 and June 1989 in Northern Italy. Onehundred fifty-seven male cases of oral cavity cancer, 134 of pharyngealcancer, 162 of laryngeal cancer, and 288 of esophageal cancer, and 1272male inpatients with acute conditions unrelated to tobacco and alcoholwere interviewed. Odds ratios for current smokers of cigarettes were 11.1for oral cavity, 12.9 for pharynx, 4.6 for larynx, and 3.8 for esophagus.For all 4 sites, the risk increased with increasing number of cigarettesand duration of smoking habits and, with the exception of esophagealcancer, decreased with increasing age at the start of and years sincequitting smoking. Smokers of pipes and cigars showed a more elevatedrisk of cancer of the oral cavity and esophagus than did cigarette smokers.Significantly increased risks emerged also in heavy drinkers (odds ratio>60 versus >19 drinks/week = 3.4, 3.6, 2.1, and 6.0 for oral cavity,pharynx, larynx, and esophagus, respectively), deriving predominantlyfrom wine consumption.

INTRODUCTION

In Western countries, cancers of the oral cavity, pharynx,larynx, and esophagus constitute from 2 to 15% of all cancerincidence (1). A comparison between age-adjusted mortalityrates for males in 27 countries has shown that many Europeancountries, such as France, Switzerland, Luxemburg, and Italy,have the highest rates for these tumors (2). In Italy, age-adjustedmortality (world population) rates among males are 6.2/100,000 for oral cavity and pharyngeal cancer, 6.6/100,000 forlaryngeal cancer, and 4.8/100,000 for esophageal cancer (2),with rates being almost double in the northeastern part (3).

Tobacco and alcohol have been well established in severalstudies as risk factors for upper aerodigestive tract cancers (4-12). Several investigations have dealt specifically with the tobacco and alcohol interaction in the etiology of cancers of theoral cavity, pharynx, larynx, and esophagus (8, 10-23). Although the nature of the biological interaction between these 2factors has not been definitively established, either multiplicative or additive risk models appear to be plausible.

Separating the effects of alcohol and tobacco remains, however, a difficult problem, since heavy drinkers tend to be heavysmokers and vice versa. Furthermore, very few persons whoneither drink nor smoke have been identified in the etiologicalstudies of cancers of the upper aerodigestive tract. To furtherclarify the role of alcohol and tobacco in the occurrence ofcancers of this type, we here report the data from a hospital-

Received3/28/90;accepted6/29/90.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1This study was supported by the Italian Association for Cancer Research,and the Italian League Against Tumors, Milan. Italy, and conducted within theframework of the Italian National Research Council Applied Project "Oncology"

(Contract 87.01544.44).1To whom requests for reprints should be addressed.

based case-control study undertaken in the northern part ofItaly.

MATERIALS AND METHODS

A hospital-based case-control study on tumors of the upper aero-digestive tract has been conducted since June 1986. Cases were malesbelow age 75 with a histologically confirmed diagnosis of cancer of theupper aerodigestive tract (i.e., oral cavity, larynx, pharynx, and esophagus). Cancers of the nasopharynx and the salivary glands were excluded. All cases had their diagnoses made within 6 months before thedate of interview and were drawn from 2 areas of northern Italy: (a)the western part of Friuli-Venezia Giulia region (Pordenone province);and (b) the greater Milan area, in the Lombardy region. The presentdata were collected before July 1989.

The 2 areas under study were not covered by cancer registries and,thus, it was not possible to estimate the proportion of upper aerodigestive tract cancers in relation to the total incidence rate. The studyhospitals, however, included the majority of diagnostic and therapeutical facilities available in the areas under surveillance and, therefore,the largest proportion of upper aerodigestive tract cancers will havebeen referred there. Furthermore, interviews were generally (90%)conducted within 2 months from cancer diagnosis, thus minimizinglosses caused by patient death and disability.

One hundred fifty-seven males with histologically confirmed cancerof the tongue and oral cavity (ICD-IX1 = 140, 141, 143, 144, and 145);

134 with cancer of the pharynx, junction between hypopharynx andlarynx included (ICD-IX = 146, 148, and 161.1): 162 with laryngealcancer (remaining ICD-IX = 161), and 288 with cancer of the esophagus(ICD-IX = 150) were interviewed. Males admitted for acute illnessesin the same hospitals were eligible as controls. None of these patientshad malignant tumors or any condition known to be related to alcoholand tobacco consumption. A total of 1272 male controls chosen on thebasis of area of residence and age within quinquennia were interviewed.Of these, 26% were admitted for nontraumatic orthopedic conditions(mainly low back pain and disc disorders), 25% for traumatic orthopedicconditions (mainly fractures and sprains), 19% for acute surgical conditions (included plastic surgery), 17% for eye disorders, and 13% forother illnesses (e.g., skin disorders). All study patients had their interviews during the course of their hospitalization. No next-of-kin respondents were used. Incidence of refusal to interview was about 2%for cases and 3% for controls.

Interviewers were trained to reduce variability between study areas,using the same precoded questionnaire to obtain information on so-ciodemographic factors; occupation; lifestyle, including tobacco andalcohol consumption habits; dietary habits; and past history of ear,nose, and throat diseases. All information referred to patient behavior,before the onset of symptoms of the disease that led to hospitaladmission. Information on smoking habits included smoking status(never, ex-, or current smoker), number of cigarettes smoked per daybefore the onset of symptoms, years of cigarette smoking, the age atstarting to smoke, and, for the exsmokers, years since quitting smoking.The part of the questionnaire relating to alcohol habits included thenumber of days per week that each alcohol-containing beverage (wine,

ìThe abbreviations used are: ICD-IX, International Classification of Diseases

Number: OR, odds ratio.

6502

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOBACCO. ALCOHOL. AND UPPER AERODIGESTIVE TRACT CANCERS

Table 1 Distribution of oral cavity, pharynx, larynx, and esophagus cancer cases and controls according to age and sociodemographic characteristics":Northern Italy, 1986-89

Site of cancer

Oral cavity<n=157)CharacteristicAge

(yr)<4950-5960-6970+Education

(yr)S45-78+Marital

statusNevermarriedEver

marriedOccupationClerical/professionalManual

workerFarmerNo.37495318398731'23134*347740*%243134112555201585235126Pharynx

(n=134)No.27405611257237*121223475A24*%2030428195428991265618Larynx(n =162)No.126177122510037*12150529116*%738487156223793335710Esophagus(n =288)No.4199111375915574"262628016822*%1434391320542699130628Controls(n=1272)No.3544363471351745315671141158495625931%2834271114424599141528

flTotal sample size varies with the number of cases and controls with incomplete information.* As compared with control group, difference was significant (P i 0.05).

beer, hard liquor) was consumed, the number of drinks per day beforethe onset of symptoms, and the duration of habit in years. Taking intoaccount the different alcohol concentrations, one drink correspondedto 150 ml of wine, 330 ml of beer, and 30 ml of hard liquor.

The ORs for the smoking and alcohol variables, together with theirapproximate 95% confidence intervals, were calculated accounting for;the study design variables—age and area of residence; potential con-founders—years of education, and occupation in 3 strata (i.e., professional and clerical, manual workers, and farmers); and the reciprocalconfounding effect of either alcohol or tobacco (24). These estimateswere obtained by unconditional multiple logistic regression (25). As afinal step, a multiple logistic regression model was fitted with interaction terms between alcohol and smoking, in addition to the terms listedabove. Attributable risks were computed by means of the methoddescribed by Bruzzi et al. (26).

RESULTS

Table 1 shows the distribution of cases and controls accordingto cancer site for age and sociodemographic characteristics.Years of education are inversely related to the diseases studied.A difference between cases and controls in marital status wasobserved only for oral cavity. With reference to occupation,controls tended to be more in the professional and clericalcategory, whereas cases tended to be more often farmers.

Tobacco Smoking. Table 2 gives the distribution of cases andcontrols. Very few patients described themselves as non-smokers. The adjusted ORs for current smokers of cigaretteswere 11.1 for oral cavity, 12.9 for pharynx, 4.6 for larynx, and3.8 for esophagus. These risks increased significantly withincreasing number of cigarettes smoked per day and durationof smoking habits for all cancer sites. Also, age at starting tosmoke showed a similar pattern of strong inverse relation torisk for all sites considered, except esophagus (Table 2). Amongexsmokers, those who had quit smoking for more than 10 yearsshowed ORs close to unity for oral cavity (1.1) and larynx (1.2),and somewhat above unity for pharynx and esophagus. Forsmokers of only pipe or cigars, risks more elevated than incigarette smokers were found for cancer of the oral cavity (20.7)and esophagus (6.7), whereas for the other tumors, the lack ofindividuals who smoked only pipe or cigars prevented us fromaddressing this issue.

Alcohol Drinking. The alcohol-related risks, adjusted for tobacco, are shown in Table 3. A highly significant direct trendin risk with an increasing number of drinks of wine consumedper week emerged for each cancer site. Significantly elevatedrisks, however, became apparent only in those who drank 56 ormore glasses of wine per week (about 1 liter/day), with theweakest associations emerging for laryngeal cancer (OR = 2.8versus 5.3, 4.0, and 4.9 for oral cavity, pharynx, and esophagus,respectively). ORs of 8.5, 10.9, 4.2, and 14.0, respectively, wereseen among those persons who reported drinking 84 or moreglasses of wine per week.

Beer and hard liquor were consumed much less frequentlythan wine, and significant risks for these beverages emergedonly for cancer of the eosphagus [OR = 1.8 for beer and forhard liquor (Table 3)]. Due to the positive correlation of winewith other alcoholic beverage consumption, all ORs for beerand hard liquor were somewhat reduced by allowance for wineconsumption. Total alcohol intake mostly reflected wine consumption and, in a similar way, showed for heavy drinkers (60or more drinks per week) the most elevated OR for esophagealcancer (OR = 6.0), and the lowest for laryngeal cancer (OR =2.1). The duration of an alcohol-drinking habit did not appearto be related to risks for any of the upper aerodigestive tracttumors considered here.

Smoking and Alcohol Interrelationship. The joint effect oftobacco and alcohol intake is examined in Table 4 in terms ofdistribution of cases and controls and in Table 5 ¡nterms ofcorresponding ORs. Cases of oral cavity and pharyngeal cancerare considered together; further, abstainers and light alcoholdrinkers (<35 drinks/week) are combined since the associatedORs (Table 2) appeared to be very similar. As concerns smokingstatus, it was defined in 4 categories: (a) nonsmokers; (b) light—exsmokers who quit >10 years ago, or smokers of 1-14 cigarettes/day for <30 years; (c) intermediate—15-24 cigarettes/day regardless of duration, 30-39 years duration regardless ofamount, 1-24 cigarettes/day for >40 years, or >15 cigarettes/day for <30 years; (d) heavy—smokers of >25 cigarettes/dayfor >40 years.

The risk of oral cavity and pharyngeal cancer for the highestlevels of alcohol and smoking was increased 80-fold relative to

6503

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOBACCO, ALCOHOL. AND UPPER AERODIGESTIVE TRACT CANCERS

Table 2 OR for oral cavity, pharynx, larynx, and esophagus cancers in males according to smoking habits": Northern Italy, 1986-89

Site ofcancerNo.

inoral cavity(n=157)Never

smoked'Cigaror pipesmokeronlyCigarette

smokerNo.

ofcigarettes/day£1415-2425+XÎ

trendYears

of cigarettesmoking1-2930-3940+XÃŽ

trendAge

startedsmoking25+17-24<|7xf

trend46147267942344969237454No.

inpharnvx

(n =134)20132306141283668147048No.

inlarynx

(n =162)81152256859204985208351No.

inesophagus(n =288)177264631109166731277015145Controls(n = 1272) Oralcavity28914967313396258414255300224498247120.711.15.314.314.339.32'5.914.318.043.61'9.210.013625.14'5.6-76.3''3.4-34.81.5-17.64.4-46.74.2-48.01.8-19.74.3-47.75.4-60.42.7-31.73.1-32.54.1-44.9OR*

(95% confidenceinterval)Pharynx112.98.014.217.629.38'6.415.525.546.16'7.912.816.025.53'3.1-52.91.9-34.53.4-59.34.1-74.71.5-27.43.6-66.76.0-109.91.7-36.13.1-53.23.8-67.5Larynx12.84.62

24J7.142.50'1.95.27.246.91'2.45.16.533.59'0.3-26.1"2.2-9.61.0-5.22.3-10.43.3-15.40.8-4.42.4-11.53.3-15.61.0-5.72.4-10.93.0-14.3Esophagus16.73.83.03.84.728.66'2.44.05.643.73'3.74.52.58.13'2.3-19.8''2.2-6.61.7-5.52.1-6.72.6-8.41.3-4.42.2-7.23.1-10.02.0-6.82.5-7.81

4-4 8

Years since quit smokingcigarettes10+<10

XÃŽtrend

520

1026

932

2941

197203

1.15.713.22'

0.3-5.11.6-20.8

3.711.320.09'

0.8-18.02.6-49.4

1.24.618.78'

0.4-3.32.0-10.4

2.22.57.60'

1.1-4.31.3-4.8

°Total sample size varies with the number of cases and controls with incomplete information.* Estimates from logistic regression adjusted for age, area of residence, years of education, occupation, and number of alcoholic drinks per week.c Reference category." Mantel-Haenszel estimates adjusted only for age and area of residence because of the small number of cases.

Table 3 OR for oral cavity, pharynx, larynx, and esophagus cancers in males according to alcohol drinking habits': Northern Italy, ¡986-89

Site ofcancerNo.

inoral cavitv

(n=157)Glasses

of wine/wk0-6'7-2021-3435-5556-8384+xi

trendGlasses

ofbeer/wkOf1-1314+XJ

trendGlasses

of hardliquor/wkOc1-67+XÃŽ

trendTotalno. ofdrinks/wk£19'20-3435-5960+XÎ

trendYears

of alcoholuse<30'30-3940+XÕ

trend12620276824111202691194715146365395360No.

inpharynx

(n=134)961628453094112873105113143473314054No.

inlarvnx

(n=162)321032275|101211625115123539275145225168No.

inesophagus(n =288)322557609420219264316735864541115876093116Controls(n =1272)Oral1472653962501862894617115178921926436033236621438536743311.11

9498.547.68"11

00.80.3010.70.90.6611.13.23.418.74''11.20.71.28cavity0.5-2.309-3

72.6-9.53.6-20.20.6-1.80.5-1.40.4-1.30.6-1.30.5-2.51.6-6.21.7-7.10.7-2.00.4-1.3OR*(95%

confidenceinterval)Pharynx10.71.93.110.946.44''10.50.90.4710.41.20.2410.91.53.621.66''11.10.90.160.3-1.60

9-3.71.6-6.14.7-25.30.3-1.00.5-1.50.2-0.90.8-1.80.4-2.00.8-3.11.8-7.20.6-2.10.4-1.8Larynx]0.81.02.64.217.69''11.11.51.7510.40.81.8710.81.32.17.19"11.51.20.060.5-1.40.6-171.6-4.41.6-10.60.6-2.10.8-2.50.2-0.80.5-1.30.5-1.40.8-2.11.2-3.80.8-2.70.6-2.2Esophagus11.21.94.414.066.79"11.11.85.25'10.81.89.58"11.03.16.068.37"11.10.90.2408-1.812-3018-6.96.4-30.606-1.61.2-2.80.5-1.31.3-2.60.6-1.72.0-4.73.7-10.00.7-1.70.6-1.5

°Total sample size varies with the number of cases and controls with incomplete information.* Estimates from logistic regression adjusted for age. area of residence, years of education, occupation, and smoking habits.' Reference category.

the lowest levels of both factors. For laryngeal cancer, thecombined effects of alcohol and smoking at the highest levelfor each variable increased the risk 12 times over that for thelowest levels. The effects of alcohol in nonsmokers were extremely difficult to assess since no cases of laryngeal cancer

were seen among nonsmokers at the highest level of alcoholconsumption.

For esophageal cancer, high levels of combined alcohol andcigarette consumption increased the risk 18 times over the riskfor the lowest levels of consumption. At variance with other

6504

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOBACCO, ALCOHOL, AND UPPER AERODIGEST1VE TRACT CANCERS

Table 4 Distribution of oral cavity, pharynx, larynx, and esophagus cancer casesand controls according to smoking and alcohol drinking habits': Northern Italy,

¡986-89

Table 5 ORs" for oral cavity/'pharynx, larynx, and esophagus cancers in malesaccording to smoking and alcohol drinking habits: Northern Italy, 1986-89

Alcoholintake<35

drinks/wkSmoking

status*Oral

cavityNonsmokersLightIntermediateHeavvTotalPharynxNonsmokersLightIntermediateHeavyTotalLarynxNonsmokersLightIntermediateHeavyTotalEsophagusNonsmokersLightIntermediateHeavyTotalControlsNonsmokersLightIntermediateHeavyTotalNo.232132914184275543126591147168318516229637680%121421913133203327841341763015132735435-59

drinks/wkNo.24524620327434333212SO3197814114817019219362%1334341022032522208311728540661512960+

drinks/wkNo.044710611855973043474551360684234912613211%03316401641755022142825212302410117Totaldrinks/wkNo.411120171522151001713481210931160174318536281289281614691253%377911211751258681961566132322496

°Total sample size varies with the number of cases and controls with incom

plete information.Smoking status defined in 4 categories: (a) nonsmokers; (h) light, exsmokers

who quit >10 years ago, or smokers of 1-14 cigarettes/day for <30 years: (c)intermediate, 15-24 cigarettes/day regardless of duration, 30-39 years' durationregardless of amount. 1-24 cigarettes/day for >40 years, or >15 cigarettes/dayfor <30 years; (d) heavy, smokers of 225 cigarettes/day for 240 years.

cancer sites, in cancer of the esophagus, the effect of drinking60 or more alcoholic drinks per week in nonsmokers wasslightly stronger than the effect of heavy smoking in lightdrinkers (OR = 7.9 versus 6.4).

DISCUSSION

Role of Tobacco and Alcohol. The associations of tobacco andalcohol with cancers of the upper aerodigestive tract have beenreported since the early part of the century (4, 5, 7). Not untilmore recently, however, have good estimates of risk associatedwith alcohol and tobacco been obtained (8-12). Some studieshave shown a strong dose-response relationship for each ofthese 2 substances after controlling for the exposure to theother (10, 11,27-30).

In the present study, elevated risks were found in all sites fornumber of cigarettes smoked per day, the length of smokinghabit, and early age at starting to smoke, which have also beenpreviously reported (10, 11, 27). In agreement with previousfindings (10, 27, 29, 31, 32), the risk associated with pipe andcigars is suggestive of a strong effect on the oral cavity andesophagus, and a more moderate effect on the larynx. A decreased risk for longer duration of quitting smoking observedhere is also compatible with the findings in another study (10),

AlcoholintakeSmoking

status*Oral

cavity/pharynxNonsmokersLightIntermediateHeavyTotalLarynxNonsmokersLightIntermediateHeavyTotalEsophagusNonsmokersLightIntermediateHeavyTotal<35drinks/wk|C3.110.917.6Ie1*0.94.56.1IeIe1.12.76.41'35-59

drinks/wk1.65.426.640.22.31.65.07.110.41.40.87.98.811.03.160+drinks/wk2.310.936.479.63.4-5.49.511.72.87.99.416.717.55.7Totaldrinks/wkIe3.714.125.0Ie1.05.46.7Ie2.54.06.6

" Estimates from logistic regression equation including age, area of residence,

years of education, occupation, drinks per week, and smoking habits, as appropriate.

* See Table 4, Footnote b.c Reference category.

and suggests that cancer risk among exsmokers substantiallydeclines after 10 years or more after cessation of smoking.

As regards the risks associated with different types of alcoholic beverages, wine seems to exert the strongest effect whencompared with beer and hard liquor. This result was not atvariance with studies from other countries (8, 10, 14, 27, 33,34) when levels of exposure to various alcoholic beverages aretaken into account, since wine is by far the predominant beverage consumed in Italy (10, 14, 27). Only esophageal cancerrisk seemed to be enhanced significantly by moderate consumption of alcoholic beverages other than wine. Due to the positivecorrelation with wine consumption, however, when the associations of esophageal cancer with beer and hard liquor wereadjusted for wine, they became weaker.

Some potential limitations in the methodology of this case-control study should be noted. First, the use of hospital-basedcontrols in studying the etiology of cancers that are clearlyrelated to smoking habits, such as cancers of the upper aero-digestive tract, has been widely criticized based on the fact thathospital controls tend to smoke more than does the generalpopulation (35). In the present study, however, the distributionof smoking and drinking habits in hospital controls (from whichpatients with tobacco- and alcohol-related diseases were excluded) turned out to be very similar to that of the generalpopulation of the same area (36). Furthermore, no inconsistencies in alcohol-related ORs emerged when the 4 major diagnostic groups (i.e., trauma, orthopedic, surgical, and miscellaneous conditions) were used separately.

Of, perhaps, greater concern is the potential misclassificationof exposure and confounding variables used in this study. Self-reports of smoking habits and alcohol consumption, in particular, may suffer from less-than-perfect reliability, and if suchreliability is differential between cases and controls, the direction and magnitude of residual confounding is not predictable(37). We believe that the standardization of the questionnaireand its administration across the study areas minimized thepotential for eliciting smoking and drinking habits differentiallyfrom cases and controls. Furthermore, there is no reason to

6505

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOBACCO, ALCOHOL, AND UPPER AERODIGESTIVE TRACT CANCERS

believe that preferential recall among cases could account forthe elevated ORs.

Smoking and Alcohol Interaction. In an effort to understandthe ways in which alcohol and tobacco act, several investigatorshave modeled the risk of disease using multiplicative riskmodels, additive risk models, and models that allow for anintermediate effect between additive and multiplicative (38, 39).Several studies of oral cavity and pharyngeal cancer (7, 8, 10,11, 20, 21, 40) have empirically examined the interaction between alcohol consumption and smoking. For the highest levelsof consumption for both factors, estimates of risk comparedwith the lowest levels of consumption range from 8.0 to 141.6.The present study estimate fell at the high end of this rangeand was compatible with a greatly elevated risk for heavysmokers who also drank heavily.

As concerns the combined effects of smoking and alcohol onlaryngeal cancer (7, 11, 12, 15, 31, 40-42), estimates of risk atthe highest levels of exposure for both factors ranged from 8.0to 22.1. Again, the results of our study were not inconsistentwith this set of values representing elevated risk, but with amagnitude lower than that for oral cavity and pharyngeal cancer.

As concerns esophageal cancer, fewer studies have examinedthe joint effects of tobacco and alcohol intake (6, 7, 14, 22),generally showing multiplicative effects. Using 10 g of alcoholper drink and 2 g of tobacco per cigarette, the risk reported inthis study was 17.5 based on approximately 90 g or more perday of alcohol and 25 g or more of tobacco per day. Such riskestimates fell in the range provided by previous work on thesubject (6, 14), although they were at the lowest end.

Attributable Risks and Conclusions. The interest of this workis to add further quantitative evidence toward the associationbetween alcohol and tobacco and 4 different neoplasms of theupper aerodigestive tract. From a public health viewpoint, thepresent study shows attributable risks (26) of over 75% forevery site for smoking and alcohol together. Smoking, however,showed a higher attributable risk than alcohol for cancers ofthe oral cavity (76% versus 55%), pharynx (69% versus 45%),and larynx (70% versus 26%). For cancer of the esophagus,alcohol showed a slightly higher attributable risk than smoking(52% versus 40%).

As noted by a number of investigators (10, 15, 25), theimplication of an interaction between smoking and alcohol,which appears to be, on the whole, greater than additive, is areduction in the occurrence of cancers of the upper aerodigestivetract by eliminating or moderating one or the other of thesehigh-risk behaviors. In northern Italy, where both smoking andheavy alcohol consumption are widespread (36, 43, 44), publichealth interventions to discourage smoking and heavy drinkinghave a great potential to result in gains to society in terms ofreductions in cancer morbidity and mortality.

ACKNOWLEDGMENTS

The authors wish to thank Tiziana Angelin and Derna Gerdol forinterviewing patients and Anna Redivo for editorial assistance.

REFERENCES

1. Muir, C.. Walerhouse, J.. Mack, T.. Powell. J., and Whelan, S. Cancerincidence in five continents. In: International Agency for Research on CancerScientific Publication 88, Vol. 5. Lyon. France: International Agency forResearch on Cancer. 1987.

2. Levi. F.. Maisonneuve. P., Filiberti. R.. La Vecchia, C., and Boyle. P. Cancer

6506

incidence and mortality in Europe. Med. Soc. Prev.. 34 (Suppl. 2): 1-84,1989.

). Franceschi. S.. Bidoli. E., Barra. S.. Gerdol. D.. Serraino. D., and Talamini.R. In: Atlante della Mortalità per Tumori nella Regione Friuli-VeneziaGiulia. 1980-83. Servizio di Epidemiologia. Aviano, Italy: Centro di Riferimento Oncologico, 1989.

4. Wynder, E. L.. Bross. I. L.. and Feldman, R. M. A study of etiological factorsin cancer of the mouth. Cancer (Phila.), 10: 1300-1323, 1957.

5. Schwartz, D., Denoix, P. F., and Auquera, G. Recherche des localisations ducancer associéesaux facteurs tabac et alcohol chez l'homme. Bull. Cancer,«.•336-361,1957.

6. Wynder, E. L., and Bross, 1. J. A study of etiological factors in cancer of theesophagus. Cancer (Phila.), 14: 389-413, 1961.

7. Wynder, E. L., and Stellmann, S. D. Comparative epidemiology of tobacco-related cancers. Cancer Res., 37: 4608-4622, 1977.

8. Elwood, J. M., Pearson, J. C. G., Skippen, D. H., and Jackson. S. M.Alcohol, smoking, social and occupational factors in the aetiology of cancerof the oral cavity, pharynx and larynx. Int. J. Cancer, 34:603-612. 1984.

9. Brugere, J., Guenel, P., Ledere. A., and Rodriguez, J. Differential effects oftobacco and alcohol in cancer of the larynx, pharynx, and mouth. Cancer(Phila.), 57:391-395, 1986.

10. Blot, W. J.. McLaughlin. J. K.. Winn. D. M., Austin. D. F., Greenberg, R.S.. Preston-Martin, S.. Bernstein. L.. Schoenberg, J. B., Stemhagen, A., andFraumeni, J. F., Jr. Smoking and drinking in relation to oral and pharyngealcancer. Cancer Res., 48: 3282-3287, 1988.

11. Tuyns, A. J.. Esteve. J.. Raymond. L., Berrino, F., Benhamou, E., Blanche!,F., Boffelta, P., Crossignani, P., Del Moral, A., Lehmann, W., Merletti, F.,Péquignot,G., Riboli, E., Sancho-Garnier, H., Terracini, B., Zubiri, A., andZubiri, L. Cancer of the larynx/hypopharynx, tobacco and alcohol: International Agency for Research on Cancer international case-control study inTurin and Várese(Italy). Zaragoza and Navarra (Spain). Geneva (Switzerland), and Calvados (France). Int. J. Cancer, 41:483-491, 1988.

12. Falk, R. T., Pickle, L. W., Brown, L. M., Mason, T. J., Büffler,P. A., andFraumeni. J. F., Jr. Effect of smoking and alcohol consumption on laryngealcancer risk in coastal Texas. Cancer Res., 49:4024-4029. 1989.

13. Rothman. K. J.. and Keller. A. The effect of joint exposure to alcohol andtobacco on risk of cancer of the mouth and pharynx. J. Chronic Dis.. 25:711-716, 1972.

14. Tuyns, A. J.. Péquignot,G., and Jensen, O. M. Le cancer de l'oesophage enIlle-et-Vilaine en fonction des niveaux de consommation d'alcool et de tabac.Des risques qui se multiplient. Bull. Cancer, 64:45-60, 1977.

15. Flanders, W. D.. and Rothman. K. J. Interaction of alcohol and tobacco inlaryngeal cancer. Am. J. Epidemiol., 115: 371-379, 1982.

16. Nolani. P. N. Role of alcohol in cancers of the upper alimentary tract: use ofmodels in risk assessment. J. Epidemiol. Community Health. 42: 187-192,1988.

17. Walter, S. D., and Holford, T. R. Additive, multiplicative, and other modelsfor disease risk. Am. J. Epidemiol., 108: 341-346, 1978.

18. Hinds, M. W., Thomas, D. B., and O'Reilly. H. P. Asbestos, dental X-rays,

tobacco, and alcohol in the epidemiology of laryngeal cancer. Cancer (Phila.),44: 1114-1120, 1979.

19. Herity, B., Moriarty. M.. Bourke, G. J.. and Daly. L. A case-control studyof head and neck cancer in the Republic of Ireland. Br. J. Cancer, 43: 177-182, 1981.

20. Franco, E. L., Kowalski. L. P., Oliveir, B. V., Curado, M. P.. Pereira, R. N..Silva, M. E., Fava, A. S., and Turioni, H. Risk factors for oral cancer inBrazil: a case-control study. Int. J. Cancer, 43: 992-1000, 1989.

21. Graham, S., Dayal, H., Rohrer, T., Swanson, M., Sultz, H., Shedd, D., andFischman, S. Dentition, diet, tobacco, and alcohol in the epidemiology oforal cancer. J. Nati. Cancer Inst.. 59: 1611-1618, 1977.

22. Feldman, J. G., and Boxer, P. Relationship of drinking to head and neckcancer. Prev. Med.. 8: 507-519. 1979.

23. Sankaranarayanan, R., Duffy, S.. Day, N. E., Nair, M. K., Padmakumary,G. A case-control investigation of cancer of the oral tongue and the floor ofthe mouth in southern India. Int. J. Cancer, «.-617-621, 1989.

24. Breslow. N. E., and Day, N. E. Statistical methods in cancer research. Vol.I—theanalysis of case-control studies. International Agency for Research onCancer. Scientific Publication 32, pp. 85-159. Lyon, France: InternationalAgency for Research on Cancer, 1980.

25. Egret. Reference Manual. Statistic and Epidemiology Research Corporationand Cytel Software Corporation. Seattle, WA, 1985-1990.

26. Bruzzi. P.. Green, S. B., Byar, D. P., Bunion. L. A., and Schairer, C.Estimating the population attributable risk for multiple risk factors usingcase-control data. Am. J. Epidemiol., 122: 904-914, 1985.

27. Merletti, F., Boffetta, P., Ciccone, G., Mashberg, A., and Terracini, B. Roleof tobacco and alcoholic beverages in the etiology of cancer of the oral cavity/oropharynx in Torino, Italy. Cancer Res.. 49: 4919-4924, 1989.

28. La Vecchia, C., Liati, P., Decarli, A., Negrello, I., and Franceschi, S. Taryields of cigarettes and the risk of oesophageal cancer. Int. J. Cancer, 38:381-385, 1986.

29. La Vecchia. C., and Negri. E. The role of alcohol in oesophageal cancer innon-smokers, and of tobacco in non-drinkers. Int. J. Cancer. 43: 784-785,1989.

30. Ferraron!. M., Negri, E., La Vecchia, C., D'Avanzo, B., and Franceschi, S.

Socioeconomic indicators, tobacco and alcohol in the aetiology of digestivetract neoplasms. Int. J. Epidemiol.. 18: 556-562, 1989.

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

TOBACCO, ALCOHOL, AND UPPER AERODIGESTIVE TRACT CANCERS

31. Wynder, E. L., Covey, L. S., Mabuchi, K., and Mushinski, M. Environmentalfactors in cancer of the larynx. A second look. Cancer (Phila.), 38: 1591-1601, 1976.

32. Tuyns, A. J., and Estève,J. Pipe, commercial and hand-rolled cigarettesmoking in oesophageal cancer. Int. J. Epidemiol., 12: 110-113, 1983.

33. Tuyns, A. J., Pèquignot, <... and Abbatucci, J. S. Oesophageal cancer andalcohol consumption: importance of type of beverage. Int. J. Cancer. 23:443-447, 1979.

34. Morris Brown, L.. Blot. W. J., Shuman, S. H., Smith, V. M., Ershow, A. G..Marks, R. I)., and Fraumeni, J. F., Jr. Environmental factors and high riskof esophageal cancer among men in Coastal South California. J. Nati. CancerInst., 80: 1620-1625. 1988.

35. Tuyns, A. J., Jensen, O. M., and Pèquignot, G. Le choix difficile d'u bon

groupe de témoinsdans une enquêterétrospective.Rev. Epidém.SantéPubi..¿5:67-84, 1977.

36. Istituto Centrale di Statistica. Indagine statistica sulle condizioni di salutedella popolazione e sul ricorso ai servizi sanitari. Note e relazioni anno 1986,N. 1. Rome, Italy: Instituto Centrale di Statistica. 1986.

37. Greenland, S. The effect of misclassification in the presence of covariates.Am. J. Epidemici., ¡12:564-569. 1980.

38. Breslow, N. E., and Day, N. E. Statistical methods in cancer research. Vol.1. The analysis of case-control studies, scientific publication 32. Lyon.France: International Agency for Research on Cancer, 213-221, 1980.

39. Breslow, N. E., and Storer, B. E. General relative risk function for case-control studies. Am. J. Epidemiol., 122: 149-162, 1985.

40. Rothman, K. J.. Cann, C. L, and Fried, M. P. Carcinogenicity of dark liquor.Am. J. Public Health. 79: 1516-1520. 1989.

41. Olsen, J.. Sabreo, S., and Fasting, U. Interaction of alcohol and tobacco asrisk factors in cancer of the laryngeal region. J. Epidemiol. CommunityHealth. 39: 165-168. 1985.

42. Waller, S. 1).. and Iwane, M. R.: "Interaction of alcohol and tobacco inlaryngeal cancer" (letter). Am. J. Epidemiol., 117:639-641, 1983.

43. Menino. F., Merletti, F., Zubiri, A., Del Moral, A.. Raymond, L., Estève,J.,and Tuyns, A. J. A comparative study of smoking, drinking and dietaryhabits in population samples in France, Italy, Spain and Switzerland. II.Tobacco smoking. Rev. Epidém.Sante Pubi., 36: 166-176, 1988.

44. Pèquignot, G., Crosignani, P., Terracini, B., Ascunce, N., Zubiri, A., Raymond. L., Estève,J., and Tuyns, A. J. A comparative study of smoking,drinking and dietary habits in population samples in France, Italy, Spain andSwitzerland. HI. Consumption of alcohol. Rev. Epidém.SantéPubi.. 36:177-185. 1988.

6507

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1990;50:6502-6507. Cancer Res   Silvia Franceschi, Renato Talamini, Salvatore Barra, et al.   Pharynx, Larynx, and Esophagus in Northern ItalySmoking and Drinking in Relation to Cancers of the Oral Cavity,

  Updated version

  http://cancerres.aacrjournals.org/content/50/20/6502

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/50/20/6502To request permission to re-use all or part of this article, use this link

on July 28, 2018. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from