Dr. Bernard Fox has served as the mentor and unbiased monitor of psychological, social, and behavioral research in cancer as it has evolved over the past 20 years. The thoughtful review in this issue of the current status of this research is extremely valuable for oncologists who must deal with patients' concerns as to whether their personality, emotions, or recent stresses caused their cancer or its progression. Media reports of psychological studies proposing new evidence of mind-body relationships and cancer are read by frightened patients who are trying to make some meaning out of their plight. The physician who has read Dr. Fox's article can provide a strong antidote to patients' inappropriate assumptions that they have somehow caused their cancer.
The review provides a critique of research in the major areas in which the most intensive work has been done on possible psychosocial influences on cancer morbidity and mortality. In several areas, Dr. Fox feels that research indicates with certainty that there is no association. First, grief, depression, and depressive symptoms are not predictors of cancer or of its progression. Likewise, stress in humans and a history of psychiatric disorders do not predict cancer or its progression. Social support and social ties, which are inversely related to overall age-adjusted mortality, are less impressively associated in cancer, although two epidemiologic studies suggested that men were more vulnerable than women to social isolation. The data regarding social support in patients with cardiovascular disease are far more impressive. Personality traits and affective states, such as anger, taken on measure, also do not appear to be important factors.
Two recent studies cited, by Fawzy and Spiegel and their colleagues, of psychosocial and psychoeducational interventions showed positive effects on quality of life and survival. These have caused a major new interest in the field. Numbers of patients randomized in both studies were small, and effect on survival must be replicated.
Overall, while Dr. Fox feels one cannot rule out that psychological factors may be an influence in some patients and in some cancers at some times, the data do not support their being a major factor, or one that can be predicted by the current state of information.
Influence of Social Environment
Dr. Fox did not review the influence of sociodemographic factors on cancer risk and mortality. In a Cancer and Leukemia Group B study by Cella and colleagues , almost 1,000 patients were studied for survival in five protocols involving treatment for several sites of cancer. When known predictor variables were controlled in each protocol, patients with less education and income had poorer survival rates than those who were more affluent, even when receiving the identical treatment by protocol.
A recent review by Adler and coworkers reviewed socioeconomic status and health . They found that a stepwise gradient exists by social class, with mortality increasing by each lower social class and income level. These data suggest that the issue is not just that a break at the poverty line indicates poorer services and access, but that some other higher order factor or factors could account for the gradient by social class. Research in this area is important.
While we await the accumulation of data from more studies that include psychological and social variables, the data are unquestioned that psychological factors are strong determinants of behaviors that predispose to cancer, such as tobacco use. Participation in cancer screening and early diagnosis impact upon outcome. Psychosocial interventions have a positive effect on all those behaviors, and on the quality of life of individuals going through the experience of cancer. Oncologists should encourage their patients to seek psychological support as a part of their general medical care. They should be discouraged from the view that counseling or psychotherapy suggests that one is "weak or unable to cope." Cancer is an emotional roller coaster that can shake the equanimity of the strongest individual.
Several recent papers, derived from multicenter clinical trials that have collected quality-of-life data from patients' own perspectives, have shown that patients' estimates of their level of positive well-being and quality of life early in their course was later associated with longer survival, even when controlled for predictor variables [3-5]. These data are beginning to give us more insight into responses to illness at each stage. The data may reveal patients' perceptions of their status that may not be obvious to the clinician. Importantly, quality-of-life data begin to factor in quality of survival along with length of survival, which provides a new dimension to the evaluation of new therapies.