onset of cancer are inconsistent; results range from no association to a strong association (Fox, 1989; Petticrew et al., 1999; Turner-Cobb et al., 2001; Duijts et al., 2003; Heffner et al., 2003; Walker et al., 2005). These conflicting findings are due in no small part to methodological limitations of this work. Some of these limitations have to do with the measurement of biological processes; newer studies are finding more linkages between stress and biological processes that may serve as mechanisms in tumor development and growth. Other limitations derive from problems in the measurement of exposure to stress and of disease outcomes. Because the incidence literature is based primarily on measures of stressful life events, associations could be obscured by the fact that those who can cope effectively with such events are less subject to disease (Eysenck, 1988; Giese-Davis and Spiegel, 2003). On the other hand, most cancers develop over many years and are diagnosed only after developing for 2–30 years, arguing against an association between recent stressful events and the onset of cancer (National Cancer Institute, 2007).

It is generally accepted that stress is more likely to influence the progression and recurrence of cancer than the initial onset of the disease (Thaker et al., 2007). This assumption is based largely on evidence that stress and depression can influence immunocompetence, and that the immune system plays an important role in tumor surveillance and growth (Cohen and Rabin, 1988; Anderson et al., 1994; Turner-Cobb et al., 2001). Yet even research in these areas has produced inconsistent results (Cohen and Herbert, 1996; Giese-Davis and Spiegel, 2003; Walker et al., 2005). The lack of impressive data on psychological stress and depression as risks for the onset, progression, or recurrence of cancer is at least partly attributable to the practical difficulties of designing and implementing adequate studies. For example, in the interest of maximizing power, studies frequently combine multiple types of cancers. Such an approach makes it difficult to interpret results, as it is likely that stress may influence the development of some types of tumors (e.g., those caused by viruses or subject to endocrine regulation) but not others. Despite the less clear evidence to date on the effect of stress on cancer, growing knowledge about the effects of stress on body function—in particular on the functioning of the immune system—adds to suspicions about the potential adverse effects of stress on the progression of some types of cancer.

Although epidemiologic studies conducted to date are inconclusive about the effects of stress on the development and progression of cancer, evidence emerging from the science of psychoneuroimmunology—the study