Most laypeople, physicians, and other health professionals have a very different view of the relationship between personal behaviors and health and illness today than they did a century ago. At the beginning of the twentieth century, most people believed that their health was a matter of concern only when they were sick. By the end of the century, most people accepted the statistical evidence that specific behaviors and characteristics of healthy persons, called “risk factors,” can increase the probability of developing disease, especially chronic disease. Most people also believe that making appropriate changes in their lifestyles, a concept associated with risk factors, can reduce the probability of the occurrence of disease. To facilitate this process, they expect health professionals, agencies concerned with public health, and the media to inform and educate the public about risk factors.

The acceptance of risk factors has produced changes in public health and medicine as profound as those that resulted from bacteriology and the germ theory of disease. The changes have been most evident in coronary heart disease, which was the major cause of death in the twentieth century and is the focus of this study. Yet the impact of the risk factor has been much more uneven than the germ theory. The risk factor concept has been controversial because of its statistical methodology, its multifactorial concept of disease etiology, and its effect on the economic interests of commercial, professional, and health organizations.

This study endeavors to explain in nontechnical language how one of the greatest revolutions in the understanding of health and disease could have produced such mixed outcomes. Although risk factors are a statistical concept, I have avoided all discussion of the mathematics involved. I have also refrained from using statistical terminology, although this has occasionally necessitated lengthier descriptions. Readers familiar with risk factor research will find that I have made little use of a popular statistical tool, meta-analyses. I believe that meta-analyses are based on the erroneous premise that methodological flaws in individual studies cancel each other when studies are combined. The history of the use of statistics in the social sciences, and I believe in medicine, provides convincing evidence to reject this assumption. Instead I have looked for agreement among studies that used the most rigorous methodologies.