EDITORIAL From epidemiology to cancer prevention: implications for the 21st Century Graham A. Colditz Received: 16 January 2007 / Accepted: 16 January 2007 / Published online: 30 January 2007 Ó Springer Science+Business Media B.V. 2007 Introduction As I move to Washington University School of Medi- cine and the Alvin J. Siteman Cancer Center I con- clude my term as editor-in-chief of the journal. Over the past 8 years the journal has increased in number of issues and pages published per year, number of sub- mission, and the scope of topics covered. The increase in articles reporting prevention interventions, risk communication, and related topics, reflects in part, the growing program of research funded by the US National Cancer Institute. As Dr. Giovannucci takes over as the Editor-in-Chief, bringing his extraordinary wealth of experience in nutrition and cancer to the leadership of the Journal, I will join the growing team of Associate Editors to work with him in the coming years. In the essay that follows, I reflect on the direc- tion our field is taking. I also identify some of the challenges that we must address particularly in cancer epidemiology and cancer prevention more generally. To place this in context, we might consider that the emergence of the ‘‘genetic revolution’’ in medicine and public health provides hope and a distraction from the underlying mission of disease prevention, health pro- motion, and the reduction in the burden of cancer. Doll and Peto noted that uncovering the mechanism for an association is less important from a public health per- spective than is knowing that the lifestyle factor causes cancer [1]. What is the mission of public health? The Institute of Medicine, National Academy of Sciences, defines the mission of pubic health as ‘‘fulfilling society’s interest in assuring conditions in which people can be healthy’’ [2]. To meet this goal, we might focus on the whole population or on individuals within the population at risk of cancer and other major chronic diseases. Rose emphasized the need for population-wide prevention strategies if we are to meet the goal of maximizing the health of the population [3]. He drew on his experience in cardiovascular epidemiology and emphasized that the majority of cases arise from among those at aver- age risk. The contrast between ‘‘population-wide’’ approaches and those focused on individuals who might be classified as ‘‘high risk’’ is an issue I will return to when we consider how epidemiology has evolved to focus on genetic and other markers of risk (risk factors) in the second half of the 20th Century. Where are we today? As Rockhill has written that risk factor findings in public health are couched in probabilistic language and in fact call for aggregate level policies [4]. We obtain average risk estimates for populations from epidemi- ologic studies. How should we use these population estimates? To inform or frame policy (or behaviors) at the population level? To identify high-risk individuals, G. A. Colditz (&) Alvin J. Siteman Cancer Center, Washington University School of Medicine and Barnes-Jewish Hospital, St. Louis, MO 63110, USA e-mail: colditzg@wustl.edu G. A. Colditz School of Population Health, University of Queensland, Brisbane, Australia 123 Cancer Causes Control (2007) 18:117–123 DOI 10.1007/s10552-007-0117-z