"It will have to be large...in order to be psychologically impressive."
Most of the studies that have "indicted" cigarettes have been studies that have demonstrated associations or relationships...This method of analysis has almost always been by amount of smoking, not by amount or frequency of anything else.
While most of these studies that have indicted cigarettes have carefully avoided saying that the association or relationships between smoking and health proves that cigarette smoking causes premature death...nevertheless, the effect of the associative data or relationship data is to imply that cigarettes are causative.
However, as a result of many years of work in the tobacco field, some of it for the Tobacco Institute, we have developed a different thesis that we think merits testing. We have observed a number of other relationships or associations that correlate highly with cigarette smoking and that in our judgment strongly support the "constitutional" theory--the theory that there are physiological/psychological types of people who are inclined towards excesses and who tend to burn themselves out early. It is ou[r] belief that a study that identified types of people who were and were not inclined towards excesses (assuming the hypothesis is true) and that showed what we suspect to be the many other correlates of "excessive behavior" --in addition to cigarette smoking, would serve to put the relationship of smoking and health into perspective. The many other correlates would tend to take the relationship of smoking and health out of the cause and effect area and reestablish it as a symptom--one of many symptoms--of a type of person who is a hard driver and is inclined towards excess in general.
...It is our thesis that if a proper study were done, a type of person who is inclined towards excesses could be identified and described and that cigarette smoking would merely be one of many "symptoms" that would identify such a person. If true, such a study would: a) help to establish the smoking and health data as associative, not necessarily causative; and b) suggest the desirability of future research, medical and otherwise, on the constitutional theory--the kinds of people that are inclined towards excesses, and ultimately whether the excesses are merely descriptive or may be contributory, if not directly causative...
...The Surgeon General's study analyzed the incidence of disease...in terms of smoking. We would not make an analysis of all data [illegible] in terms of smoking. We would tend to analyze everything by everything. While we would show the amount of sleep normally gotten according to whether people are non-smokers, ex-smokers, light smokers or heavy smokers, we would also show the amount of sleep people get according to whether they are non-anchovy-eaters, occasional anchovy eaters, moderate anchovy eaters or heavy anchovy eaters. We should show the amount of tension people report living under according to their smoking habits, but we would also show it according to their coffee drinking habits and their candy bar eating habits...
...We propose a large sample for the conduct of this study.... It will have to be large...in order to be psychologically impressive.
END OF DOCUMENT