Tobacco Documents Online

C C December 20, 198'3 MEMORANDUM TO: W.T. Hoyt CC: WDH RFG FROMi: Hilda Zahn SCS RCH SUBJECT: American Publ'ic Health Association Dallas, Nov. 15-17, 1983 The subject of smoking seemed to draw, more than the usual share of attention at this year's APHA meeting.. Aside from the expected offerings on ed'ucation, cessation, smoking in pregnancy, and fire-safe cigarettes, there was more emphasis on public health aspects. Whether all this attention, was truly a legitimate part of the meeting theme of "Science and Social Action for Health and Peace" didn't matter. There are few better forums than APHA's annual gathering for beating the drums against tobacco. One session, on "Issues in Smoking Research.,"' only rated a very small room, but it was jammed with an irate auldience, irate because "the biggest issue" in public health was given such a small meeting place. The second in command at the National Institute of Drug Abuse (see item No. 1 below), suggested that whoever in APHA assigned the roomiwas a smoker. A series of papers on smoking habits of Hispanics also drew a: large audience. Some of the information that was forthcoming noted that past data about lower lung canicer mortality in Hispanics may have beenimisleading; increased lung cancer rates can be expected in this particular population. Overall attendance was considerably below that of recent years. A number of people said travel budgets were tight and that most public health professionals are in areas quite distant from Dallas. Regardless, those present exhibited the enthusilasmithat's oftenia hallmark of APHA sessions. Press attendance was not extensive smoking papers seemed limited, at least USA Today that briefly touched on a few during the meeting and coverage of nationally, to a the many story in relatively minor items. Great American Smokeout antismoking material from, meeting iin Anaheim, CA. That story, in fact, also referredi to the (of the the American Cancer Society), and some concurrent American Heart Association r The APHA itself got inito the act (again) as its Governing Council adopted a resolution urging a "smoking-free" society in U.S. by the year 2000. The full text of thie resolutionlis given the last itemiof this memo. It was learned Health Health I that the 1980 National Interview Survey, conducted by the National Ceniter Statistics, should be out shortly. It will contain eonard UiC RELATIONS COUNSEL ~ PUM9 ss«~ii for data i onm W N 13 LINCOLN ROAD •P.O. BOX 223 •GREAT NECK, N.Y. 11022 •(516)482-5715

2. smoking behavior trends, some of which were described by various speakers at Dallas. The highlights: 1. "Tobacco addiction and tobacco moirtality: implication for death certification"' -- R.T. RAVENHOLT, assistant director, Nation- al Institute on Drug Abuse, Rockville, MD. The title of Ravenholt's paper was a little misleading. Though he discussed addiction in his prepared text, he spoke mainly of the need to certify smoking,as an underlying cause of death on death certificates. In 1979 there were more than 3501,000 deaths in the U.S. due to smoking, yet less than 1% of death certificates mentioned smoking. The anatomic diagnosis was enitered, but the underlying cause of death almost never was. Ravenholt called this an "extraordinary omission." Until now, few physicians or pathologists have followed the lead of CHARLES TATE, a certifying,physician in Miami, and JOSEPH DAVIS, a Florida medical examiner. Raveniholt said he has,spokenito leaders of the District of Columbia Thoracic Society, but got a lukewarm reception. He quoted a pathologist as telling him: "Well, if I started to certify these deaths due to smoking, then I'd have to certify them due to all kinds of other environmental factors." Charting the life-timie smoking experience of patients by physicians or hospitals will provide a better basis for estimating the role of tobacco as the und'erlying or contributing cause of death, Ravenholt continued. Many physicians aren't adequately aware of the fundamental importance of death certification in determining public health action. Ravenholt said he'd spoken recently to MELVIN BELLI, the well- known San Francisco litigator. He said Belli was eager to sue. tobacco companies (again), if he can get good cases, but he needs someone to certify that the personiactually died of smoking. Physicians must overcome the misapprehension that they need "gold-plated proof" of causality before daring,tolmention smoking on death certificates, Ravenholt said. 2. "Patterns of smoking blehavior" -- LEON~ARD SCHUiMAN', Minneapolis. D3734fi33 The purpose of this presen`tation by Schuman, long involved in antismoking activities, was to update patterns of smoking behavior by citing trends from various National Health Interview Surveys conducted by the National Center for Hiealth,Statistics. He included the latest survey, done in 1980, which is to be released soon. Some interesting trends: A 42% increase iniformer smoker prev- alence over a 15-year period for all male adults. The highest .

C 3. cessation rates were in those 65 years and older. Black males showed a 67% increase in cessation, indicating,greater quitting rates than weee seen among wehite males. For females, the proportion of former smokers was greatest in those aged 29-34 years in 1965, but in 1980 it was for those aged 35-44. This is probably a cohort effect, Schuman said. (These age groups encompass the child-bearing period.) The age distribution of both current and former female smokers was quite similar for whites and: blacks. From 19701-80, the increase in the amount smoked daily was greater for females (13%) than for males (5%). Schuman questioned whether this increase might not be related to "unsatisfied nicotine craving" induced by the more popular low-yield cigarettes. This appears likely since the proportion of smokers consuming 4:0 or more cigarettes daily increased from 11.4% in 197&to 16.8% in 1980. Male quitters smoked about 3 cigarettes more daily than cur- rent smokers prior to cessation. This wasn't true for female quitters. However, there was a slight tendency for older smokers of both sexes to have smoked' a greater number of cigarettes prior to quitting. No association was apparent between the amount smoked and levels of education. The 1980 survey provides a wealth of data, especially for the behavioral scientist, to evaluate cessation education and behavior mo ification. Recidivism is still troublesome, but at least there has been an increase in attempts by current smokers to quit. Schuman said he was heartened by findings from the 1978 Roper poll on the attitudes of Americans toward smoking and the tobacco industry. 3. "Public health policy and smoking: implications for re- searchi" -- CAROL SUSSMAN, Washington, DC. Sussman, a press officer with the Nation~al Institute on Drug Abuse, reviewed the majjor factors that influence public policy and legislative efforts in relation to tobaccoluse. Commenting on the importance of the economic contributions of the tobacco industry, she quoted figures from.a 19811 "Tobacco Situation:" publication of the Department of Agriculture. On an annual basis, some tobacco-related activities represented more than 2 million jobs, $30-billion in wages and earnings, $15-billion in capital investments, and $22-billiioni in tax revenues to state, federal and other agencies, etc. _ - 03734634 The political influence of tobacco growers is considerable because of: the economic value of tobacco as a cash crop, the concentrations of tobacco farming in, a geopolitical area, and thee relatively significant numbers of fa:rmers and their employees. .

4. The strength and influence of tobacco manufacturers and their service providers can be attributed to tobacco's contributions to the economy an-d~the dependency on tobacco as aniindiustry by large numbers of people. These factors have served the industry by de- veloping and maintaining tobacco interest groups iniCongress. Smokers, 54 million strong, affect legislationiand research funding efforts. Even nonsmokers have been known to lobby against any attempts by the government to regulate personal' behavior, though this is sometimes done with tobacco industry money. -- Congress is now reviewing proposals for more smoking and health legislation. Sussman stressed the importance of the coali- tion of voluntary health agencies to further legislative efforts. 4. "The eplidemiologic basis of public health policy: preven- tion potentials for leading causes of deathi. Heart disease." -- LEWIS KULLER, Pittsburgh. Focusing on coronary artery disease (CAD, Kuller, a major figure in the MRFIT'projiect, summarized past and present data showing,a 40% decline in coronary disease mortality since 1950. This decline has occurred in all races and' in both sexes, withithe greater decline appearing in the younger groups. CAD distribution is variable both within and between popula- tions and is primarily determined by combinations of risk factors that include high blood pressure, specific lipoprotein metabolism (especially the levels of highidensity lipoprotein, or HDL, chol- esterol and low density lipoprotein cholesterol), exercise behav- ior, and, "most important, a very strong genetic component,"' Kuller said. In addition to age, the most important determinant of coronary heart disease (CHD)' is male sex. Women have lower CHD mortality and morbidity than do men, even after adjustment is made for the mlajior risk factors. The reasons for this aren't fully understood. Only the presence of diabetes mellitus, preexisting CAD or early surgi- cal castration eliminates the sex differences. Use of oral contraceptives is associated with a small,but significant, increased risk of coronary thrombosis among older premenopausal women and cigarette smokers. Several recent epidemiological studies have strongly suggested~ that estrogen therapy for post-menopausal women is associated with a substantial decrease in1CHD mortality, perhaps to a third. Estro- gen raises HDL cholesterol leve2s and is probably dose-related. This has very important implications. The potential value of es'tro- gen in reducing risk would substantially outweigh the potential hazard of uterine cancer that has been linked withithis hormone. 03734635 Cigarette smoking has decreased in recent years and this may have conitributed to the decline in coronary disease mortality, especially in those with a reduced fat diet. The primary effect of

C 5. smoking may be related to the acute event rather than the initial development of coronary atherosclerosis, and may be secondary to elevated carboxyhemoglobin levels or to the effects of nicotine. Smoking of filter-tip cigarettes doesn't appear to substantially reduce coronary disease risk. Smoking cessation is associated with a red'uction in CAD risk, generally within a relatively short time after quitting. Several studies of heart attack patients have shown that quitting smoking may reduce the subseq,uent risk of coronary mortality very substantially. Smoking cessation as a secondary prevention following a heart attack is probably as important a therapeutic effect as any type of drug, surgery or exercise rehabilitation program. - 5. ",The epidemiological basis of public health policy: prevention potentials for leadd~ng causes of death. Cancer." ABRAHAM LILIEN,FELD, Baltimore. As background~for developing public policy for cancer control, Liliienifeld, first gave some data from epidemiological studies on cancer types and sites. Environmental or exogenous agents, which he said are respon- sible for 80-g0% of cancer deaths, can be loosely classified into three categories: chemical, physical and biological. Different agents may produce cancer at the same site while a certain agent may produce cancer at different sites. For some chemical agents, carcinogenic activity may result from their metabolites since the agent iitself may not be a:carcinogen. No threshold level of dosage below which aniagent is noncarcinogenic has been definitely ascer- ta:ined, probably because it's difficult to show an effect at low dosage levels. The incubation period for cancer ranges from 5-30 years. With the extraordinary increase in the industrial productionlof chemi- ca:ls in'th.e last 20 years, future rates.for cancer morbidity and mortality willl be affected. There's considerable difference of opinion on estimates of the percentage of current mortality attributable to occupational ex- posure. The range is from 4-15%. Chemical agents an~d tobacco use add upito 30%. Some hormones are diefinitelly incriminated as carcinogens, for example, estrogen in endbmetriah cancer. However, the numbers are small -- only i -2%. ~ 03'734636 Diet may contain or form a carcinogen, such as a nlitrosamine or aflatoxin. Estimates of cancer deaths attribuitable to diet vary wildly among investigators, ranging from,20-70%'. Yet dietary fac- tors also may act as anticarcinlogenis. There's really no knowledge of a prudent diet regarding cancer, Liliienfeld said~, adding that it's foolhardy to assign a liarge role to dietary factors.

6. Alcohol is responsible for 3'% of cancer deathis. It also acts synergistically with tobacco in oral cavity and esophageal cancer. Phiysical agents like ultraviolet light and radiation account for 3% of canicers. There's strongly suggestive evid~ence that implicates several viruses as etiological factors in about 10% of certain cancers: Herpes virus for cervical canicer, Epstein-Barr virus in niaso- phiaryngeal cancer and perhaps hepatitis B& virus in liver cancer. "Tobacco use stands out as a major, well-substantiated cause of 30% of cancer d'eaths," Lilienfeld said. It has beeniestimated that lung cancer wi~ll be the most common cause of cancer deathiin women by 1986 or 1988. Hie suggested the establishment of'sta:te counci,ls on environ- mental cancer by state legislatures. These couincils should co- ordinate alil cancer activities, maintain a registry of toxic and carcinogenic substances, collect and evaluate all pertinenit dalta, etc. All this would help research andpolicy making. Educational programs are alsoidesirable, and a percenitage of tobacco taxes should be allocatedi to pay for them. Tobacco sales should be banned in all publicly owned and operated buildings. During discussion, Lilienfeld said some observers feel that the carcinogens ini cigarettes are in the tar and that cigarettes with less tar will result in a decrease inithe increase of lung cancer.mortality. Another question drew this response: "I think you can, primarily prevent breast cancer from smoking." 6. "Smoking behavior among Hispanics" -- ALFRED MARCUS, Los Angeles. Surveys have shown Latinos have a:lower risk for lung cancer and other smoking-linked cancers than Anglo Americans (AA) andi blacks because of lower smoking rates. Believing the evidence iin- complete and perhaps misleading, Marcus offered his own review. He said the overall lower smoking rate for La:tinos was more reflective of the behavior of womien than men. His analysis of a11979-801nationa:l survey of smokiing rates for AAs, bl'acks and Laitinos by sex showed Latino womerr to have rates approximately 30% lower than AA or black women. Latino meni, on the other hand, had rates as hi,gh or higher than AA men. This differ- enitial held true for Puerto Ridans, Mexicans, Mexican-Americans, other Latin-Americans and other Spanish. 03'73453'7 Analysis of smoking,rates fromia household survey of 11,250 Mexican-American (MA) women in four states bord'ering Mexico dis- closedi that they smoked less than AA women in all age groups between 15-44 years and at all educational levels. They begin smoking later and smoke less per day. Lung cancer rates were lower

4F C 7. r for the MA women and should remain relatively low. Hlowever, it was found th~at the difference between AA and MA women was considerabl.y reduced, in the 15-19-year group. Data from California and Texas also showed lower smoking rates for Latino women compared to AAs, but the rates for Latino mien were at least as high or higher than for AA men. Recent reports from the San Antonio Heart Study support these findings. ('This study is . looking,into diabetes andicardiovascular risk factors amiong MAs and AAs aged 25-65 years living in three neighborhoods -- low-, mididle- and high-income.), As for Latiino youths, high school students in southern Arizona mirrored' the trend for adults. In New York City, cigarette use among Spanish high school students was about 13% lower than for whites or blacks (there was no da:ta by sex). "Current evidence suggests Latino males smoke as frequently as their white Anglo counterparts. We may soon witness a virtual epidemic of lung canicer and other cigarette-related diseases within the Latino community," Marcus said.. 7. "Cigarette u~se and occupational exposures among a general population sample of New Mexico's Anglos and Hispanics" -- CHARLES HUMBLE, Albuquerque, NM. He gave data on the prevalence of exposure to lung cancer risk factors in New Mexico's general population: The da:ta were from ani ongoing population-based, case-control study of lung cancer in Hispanics and AAs. The study includes information on smoking histories, residential andioccupational histories, anid consumption of foods rich in vitamin A. Young male Hispanics smoked more in comparison to AAs, but female Hispanics smoked less thani half their AA counterparts. This increased smoking among younger Spanish males andiyounger AA females suggests "we will see large increases in the rates of lung cancer. Inifact, we already see this in our provisional rate cal- culations. " There was no difference in the percentage of AAs and Hispanics ever employed in jobs having exposures associated with lung cancer, though Hispanics were twice as likely to have been exposed to asbestos and five times as likely to have been exposed to wood dust. "This higher prevalence of occupational exposure seen in the Hispanics is consistent with the socioeconomic differences between the two groups in the sample. But they're not associated with the higher prevalence of lung cancer risk in the Hispanics. This points up the greater importance of smoking in the risk factors for lung cancer," H'umble concluded. ~ U3'734638 8. "Smoking in Hispanics" -- STEVEN HAFFNER, San Antonio. While studies in the 1950s found that foreign born MA females had an excess mortality rate from lung cancer, more recent research has shown that both MA males and females have low rates of lung cancer and cigarette smoking. Since most MAs have low income, the

8. lower rate of smoking behavior in MAs than AAs might represent economic rather than cultural factors. Data from the San Anton~io Heart Study show: With rises iin socioeconomic status (SES), fewer individuals are current smokers. When there is control for SES, approximately equal proportions of MAs and AAs are current smokers; an exception appears to be that in the high-income'area fewer MA females than AA are currenit smokers. MA and AA males have equal durationiof smoking while MAA females have a shorter duration of smoking than AA females. MAs in all groups smoke a third! to two-thirds the amount of their AA counterparts. This suggests that acculturation may be more impor- tant than economic factors in explaining differences in intensity of smoking among current smokers. The amount of cigarettes smoked increased, but the duration of smoking decreased as MA males became more acculturated, even when there was adjustment for SES. The probability of quitting smoking was positively related to the degree of acculturation in MA males. 9. "Ethnicity gradients and! cancer risk in Colorado's Hispanic population"' -- DAVID SAVITZ, Deniver. This poster presentation described what was called a "first step" imexamlining a! possible lifestyle basis for Hispanic cancer risks. The findings: While there was not a strong,cancer risk gradient in relation to the ethnicity indicators overall, several specific cancer sities did vary systematically with the Hispanic concentration. The expected pattern of "'more Hispanic" risk in "more Hispanic" areas was seen for male biliary tract cancers, male. .lung cancer, male leukemias and female cervical cancer. The opposite pattern was seen for male colon cancer. No pattern was observed for stomach cancer in either sex, male prostate cancer, female coloni cancer, or female breast cancer de- spite well-established Hispanic/Ang1o differences for these sites. The lung cancer pattern was most pronounced in the higher socioeconomic areas and for oat cell carcinomas. The female cervi- cal cancer pattern was strongest in the lower socioeconomic areas. 10. "Sulfuric acid: acute and chronic human health effects" -- COLIMSOSKOLNE, Toronto. 03'734639 Sulfuric acid,, a major constituent of acid'lrain, has been thought to cau~se only acute health effects iin humarns. However, as atmospheric concen:trations of sulfuric acidihave increased, there has been a concomitant increase.in upper respiratory cancers in industrializedicountries. In particular, incidence rates of lar- yngeal cancer are rising iniexcess of 3% annuall!y. r.

Experimental animals exposed to sulfuric acid have undergone histopathologic changes in the bronchial mucosa and degenerativee changes in res'•.piratory epithelium. By inference, chronic exposure by humans could cause a series of irritations that might be inter- preted retrospectiveliy as a history of ear, nose and throat con- ditions. Such conditions have been associated an~ecd'otally with the development of upper respiratory cancer. A recent study by Soskolne and a colleague found a 13'-fold' excess risk of laryngeal cancer iin industrial workers heavily ex- posed to sulfuric acid. This substance may act as a promoter in carcinogenesis. 11. "Smoking and drinking behavior change from before to d!uring,pregnancy for mothers of live born and stilllborn infants" -- KATE PRAGER, Hyattsville, MD. Data from two 1980 national surveys of live born and stillborn infants were used to show the deleterious effects of maternal smoking and d'rinking habi~ts on the fetus. These behaviors were assessed! both before andiduring pregnancy. Race, Hispanic origin, age and education were compared in relation to these behaviors. Before pregnancy, smoking rates were higher among whites than blacks, Hispanics or other races. Rates decreased with age and education. Drinking also was more common among whites, increasing with age and education. Typically, smokers were white, younger and less educated. Drinkers were white, older and better educated. During pregnancy, mothers were much more likely to stopi drinking (30%) than smoking (18%'). The tendency to stop smoking was directly related to education, but the tendency to stop drinking was inversely related to age. Smoking and drinking in mothers of live and stillborn babies were fairly similar, though for nearly every'subgroup, the prevalence of smoking,was higher and drinking lower in mothers of the stillborn. Those who abstained from both smoking and drinking were mostly Hispanic, black and of other nonwhite races. Given the levels of maternal smoking and drinking observed in this study, the former may have been more harmful to the fetus than the latter. 12. "Maternal occupationiand industry and pregnancy outcome" -- SHARON SHILLING, Cincinnati'. 03'734sqo The 1980 National Natality and Fetal Mortality Surveys were used to identify some possible adverse pregnancy results from employment (or unemployment) in various occupations and industries Maternal characteristics such as age and smoking status also were analyzed.

10. ~ Providing some highlights from preliminary data, Shilling (she's with the National Institute of Occupational Safety and Health), said she foundi no occupation that had a disproportionately larger number of adverse pregnancy outcomes. One subgroup, sales clerks in retail trade, had a significantly smaller percentage of malformed infants. Mothers who were 30 years or older had a greater number of malformed infants and late fetal deaths regardless of their employ- ment status. A larger percentage of mothers with low birth weight children were under age 20, though they had fewer children if they were employed. Unemployed blacks had a larger proportioniof low weight births and employed blacks had a larger proportion of late fetal deaths compared to all live births. A greater percentage of mothers having low weight birthss smoked during pregnancy. More employed pregnanit mothers drank alcohol, but there was no significant difference in birth outcome. This was also true for coffee or tea use. 13. "Head and~neck cancer survival and lifestyle change" JOHN GARDNER, Salc Lake City, UT. This poster presentation described a study of 200 patientss with head and nieck canicer (oral cavity, larynx and pharynx). Thios who continued smoke had a four times greater risk of recurrent or persistent disease tharn nonsmokers and twice the risk of those who stopped smoking. Survival rates for nonsmokers were higher than for continuing smokers, and those who stopped had survival rates ap- proaching those of nonsmokers. Alcohol consumption was found to increase the risk of acquiring these cancers 3.6 times, smoking alone 5.8 times, and~ both factors increased the risk 19 times. Continued drinking didn't affect survival rates. 14. "Myocardial infarction (MI) in men under age 55 in relation to exercise" -- LYNN ROSENBERG, Cambridge, MA. A case-control study of abouit 1,500 mien aged 25-54 years with first MIs was don~e inithree New England hospitals to determine the value of exercise. Some 2,000 conitrols, being treated for infec- tions, traumas, etc., also were interviewed in the same hospitals. Data on both groups included' age, marital status, diabetes, elevated cholesterol, smoking,,alcohol use, and Type A behavior. 03734641 Exercise habits were divided into active (tennis, swimming, jogging, etc. ) and~nonactive ('golf, softball, bowling, etc. )'. It was found that men who actively exercised at least 5 hours weekly appeared to have half the risk of MI of those who:didn't exercise actively at all. Active exercise for less than 2 hours