¹ Roach's smoking habits were described in an article entitled "100 Years of Hal Roach", which appeared in The Washington Post on January 24, 1992. His death was reported in the December 13, 1992, edition of the Post. No date or cause of death was given.

² The Seychelles, which I mentioned earlier, are inhabited 99% by Blacks, and have an extremely low rate of male lung cancer. Oddly, however, African Americans, living in the U.S., are said to have a 50% greater LCDR than Caucasians, despite the fact that African Americans smoke less than their Caucasian counterparts. See, Progress against cancer, by John Carpi, American Health, v13, issue 8 (Oct., 1994).

³ The American Cancer Society estimates that in 1993, there were 153,000 deaths from "lung cancer". See World Almanac and Book of Facts, 1994 Edition. This figure is comparable with the figure in the Statistical Abstract of the United States, for deaths from lung cancer, including deaths from cancer of other "intrathoracic organs", e.g., the esophagus and the throat. I believe, therefore, that the Cancer Society estimates, like the figures given in the Statistical Abstract, tend to exaggerate the number of lung cancer deaths, because they include cancers of "other intrathoracic organs".

⁴ Report, p. 14.

⁵ Report, p. 23.

⁶ It is not close to the consumption figures given in International Smoking Statistics (ISS), a book published under the auspices of the Wolfson Institute of Preventive Medicine, which I will use as a source, later on. According to ISS, annual per capita consumption of cigarettes in the U.S. has changed relatively little between 1961, when it was 10.9 per adult per day, and 1985, when it was, allegedly, 8.8 per day, a decline of only 19.2% (ISS, at p. 453). There is something wrong with the ISS cigarette numbers. The same publication shows that annual consumption of all tobacco products in the U.S. has declined from a peak of 13.8 grams per day in 1963 to 8.4 grams in 1985 ISS, p. 453). That's a decline of almost 40%. Since the overwhelming bulk of tobacco goes into cigarettes and, according to the same source, the percentage going into cigarettes changed very little between 1965 and 1985, there is no way that a 40% drop in tobacco consumption could translate into a drop in cigarette consumption of only 20%. I suspect, but cannot prove, that the ISS figures do not properly exclude cigarettes made for export. They do not, for example, take into account the large number of cigarettes made for the U.S. domestic market, which are smuggled into Canada to evade high Canadian taxes.

⁷ To make any sense, cancer rates must always be age adjusted, to take into account the aging of the population. If a population contains a large percentage of old people, it is likely to have a high incidence of cancer, because cancer is a disease of old age. Conversely, if a population contains a high percentage of young people, it will have a low incidence of cancer because young people generally don't develop the disease. Over time, the percentage of young or old people in a country may change. Age adjustment corrects for these changes so that statistics for any particular year may be compared with those for another year, without the distortions which would otherwise result from changes in the aging of the population.

⁸ International Smoking Statistics, Oxford University Press, ISBN 0 19 262485 7 (1993), at pps. 457 and 471.

⁹ From "Facts on File", for 04/29/59.

¹⁰ In International Smoking Statistics, which I've previously cited, there is mention of a Fortune magazine poll, taken in 1935, which showed that 26% of the women in the U.S. from age 20 through age 39 smoked, and 9% of those over that age. The authors, however, state that the "age range and product are uncertain". Page 472.

¹¹ Source: Statistical Abstract of the U.S., 1993.

¹² Cited in Brownlee, "A Review of 'Smoking and Health'", 60 Journal of the American Stat. Assn., 722 (1965).

¹³ Recent medical advances in the field of emphysema leave little doubt that Reynolds' case was genetic. In an article in Grolier's Encyclopedia, Howard Buechner, M.D., explains that a significant number of people with the disease lack a gene that controls the liver's production of a protein called alpha-1 antitrypsin (AAT). This protein degrades or controls an enzyme called neutrophilelastase, produced by the white blood cells. When the enzyme is left unchecked, it destroys alveolar (lung) tissue.

¹⁴ The costs of Poor Health Habits, A RAND study, Cambridge, MA.; Harvard University Press, 1991.

¹⁵ Average state and local taxes. Source: Cigarette Taxes to Fund Health Care Reform, Congressional Research Service Report No. 94-214E.

¹⁶ "Flyer Complaints of Illness are Subject of Studies on Plane Cabin Air Quality", by Carl Quintanilla, The Wall Street Journal, July 29, 1993.

¹⁷ Brownlee. K.A. (1965), A Review of "Smoking and Health", J. Amer. Statist. Ass. 60, 722-739.

¹⁸ Considering each animal to be one experiment.

¹⁹ Source: 1964 Report, p. 83, and the references cited therein.

²⁰ Berkson, J. The statistical study of association between smoking and lung cancer. Proc. Staff Meeting, Mayo Clin 30: 319-48, 1955.

²¹ Fischer R.A.: Smoking, The Cancer Controversy, Some Attempts to Assess the Evidence, Edinburg; Oliver and Boyd, 1959.

²² Winkelby et al., "Social class disparities in risk factors for disease: Eight year prevalence patterns by level of education" Preventive Medicine: 19:1 (1990).

^22a Kitigawa & Hauser, "Differential mortality in the United States: A study in socioeconomic epidemiology", Harvard University Press (1973).

²³ A Canadian study showed that people with a high SES had a 1.9% chance of suffering a major psychological depression; people in a medium SES had a 4.5% chance, and people in the lowest SES had a 12.4% chance. Murphy, et al, "Depression and anxiety in relation to social status", Archives of General Psychiatry (1991).

²⁴ "The original case-control studies by Wynder and Graham and by Doll and Hill are still used in a famous epidemiologic exercise....where they serve as examples of what can go wrong: biased ascertainment of exposure, selection of cases and controls from different source populations, poor ascertainment of caseness, etc..." From page 427 of Invited Commentary: How Much Retropsychology?, by J.P. Vandenbroucke, Department of Clinical Epidemiology, Leiden University Hospital, American Journal of Epidemiology, Vol 133, Number 5, March 1, 1991, (pages 426-7).

²⁵ Heasman, M.A., and Lipworth, L. (1966) Accuracy of Certification of Cause of Death, Studies on Medical and Population Subjects, No. 20, HMSO, London.

²⁶ Necropsy Evidence of Detection Bias in the Diagnosis of Lung Cancer.

²⁷ Smoking and lung cancer: the problem of inferring cause (1978), J. Royal Statistical Society A 141:437-477.

²⁸ See description of Burch's papers in Eysenck, H.J., Smoking, personality and Stress, ISBN 0-387-97493-9, at page 34 (1991)

²⁹ The experimenters used a mixture of 10% cigarette smoke and 90% air. Evidently, "sham exposed" refers to 100% air.

³⁰ Toxicol App. Pharmacol (99) (1).

³¹ Cancer Res. 45 (11 Part 1).

³² European Respiratory Journal 6 (8). 1993 (1173-1180).

³³ Charlotte Roneus, however, a contributor to the alt.smokers Internet group, reminds me that Edgar Cayce, the famous psychic healer, recommended smoking four to six all-tobacco cigarettes per day as a remedy for asthma. Charlotte tells me that she has asthma, and finds that smoking cigarettes soothes the lungs and relieves the attacks. So, it's possible that the Norwegian experiments do, in fact, tell us that cigarette smoke can relieve lung irritation. While anti-smokers may find this difficult to believe, smokers, from experience, will not. Of course, when someone is ill with a cold or flu, smoking becomes unpleasant. But that is not because of smoke irritation. Rather, it's because the body becomes alkaloid when there is a fever, and the nicotine in smoke is an alkaloid.

³⁴ Environ Res 52 (1). 1990 22-23.

³⁵ Retrospective studies are studies in which the relatives of deceased persons are questioned about the deceased's smoking habits. In such studies, there is a tendency, if the deceased died from lung cancer, for the survivors to exaggerate his smoking habits. This tendency flows from what psychologists would call "suggestion". Knowing that smoking is supposed to cause lung cancer, people tend to put two and two together and get five. "Oh yes, now that you mention that he died of lung cancer, I seem to recall that Harry was a heavy smoker; I can almost see the cigarette in his lips".

³⁶ Studies Conflict on Estrogen Tie to Heart Attack, by Jerry E. Bishop, The Wall Street Journal, October 24, 1985.

³⁷ Baldness in Males and Heart Disease may be Connected, by Glenn Ruffenach, The Wall Street Journal, February 24, 1993.

³⁸ "Base Metal: Heart-Attack Study Adds to the Cautions about Iron in the Diet", The Wall Street Journal, September 8, 1992.

³⁹ Doll R, Hill A. B. Lung cancer and other causes of death in relation to smoking, Br Med J 1956;ii:1071-81.

⁴⁰ Multiple risk factor intervention trial, JAMA, 248, 1465-77 (1982).

⁴¹ Writing in the journal Circulation, in 1990, M. O. Kjelsberg commented that "Two factors appear to have contributed to this more favorable mortality trend for the SI [Special Intervention] Group: (1) a change in the diuretic protocol for SI men about 5 years after randomization, which involved replacement of [one blood pressure lowering drug with another]; and (2) a favorable effect of intervention on nonfatal cardiovascular events during the trial years. In addition, delay until the full impact of beneficial effects on mortality end points from smoking cessation and cholesterol lowering could have contributed. Circulation 82 (5) 1616-1628, emphasis supplied. Free translation: the drugs they were giving the people in the intervention group were either killing them or not saving them. They fixed the problem. There was no evidence that smoking cessation had any effect.

⁴² Mortality rates after 10-15 years for participants in the Multiple Risk Factors Intervention Trial, JAMA, 263, 1795-1801 (1990).

⁴³ Ockene, et al., "Cancer in the Multiple Risk Factor Intervention Trial", Am J. Public Health, 80 (8). 1990. 954-958.

⁴⁴ World Health Organization European Collaborative Group (1982), described by Eysenck, H.J., in king, Personality and Stress, BN 3-540-97493-8, Springer-Verlag press (1991), at p. 13.

⁴⁵ As described by Eysenck, Smoking, Personality and Stress (cited previously) at p. 13.

⁴⁶ William Morrow and Company (1991).