Content area
Full Text
REVIEW
http://www.nature.com/clinicalpractice/cardio
Web End =www.nature.com/clinicalpractice/onc
Do stress-related psychosocial factors contribute to cancer incidence and survival?
Yoichi Chida*, Mark Hamer, Jane Wardle and Andrew Steptoe
SUMMARY
SUMMARY
REVIEW CRITERIAInformation for this review was compiled by searching the MEDLINE, PsycINFO, Web of Science, and PubMed databases for articles published until 2 October 2007, and assessing reference lists from relevant reviews and articles. Electronic early-release publications were also included. The main search strategy was (cancer OR carcino* OR tumor) AND (anxiety OR coping OR depress* OR personality OR psych* OR social support OR stress) AND (longitudinal OR prospective). Criteria for inclusion were publication of articles in English language in a peer-reviewed journal, with a prospective design and investigating associations between stress and related psychosocial factors and cancer incidence or mortality. Other criteria related to the analysis are described in the text.
INTRODUCTIONThe provocative question of whether psychosocial stress is a factor in cancer etiology and survival has attracted attention for centuries. Galen observed that women with a melancholic disposition were more likely to develop cancer than their counterparts with a more sanguine disposition.1 Although cumulative evidence has identified genetic, environmental (e.g. pollution, infection, and geophysical factors), lifestyle (e.g. tobacco, alcohol, diet, and physical activity), endocrine, and socioeconomic factors as potentially increasing the risk of carcino-genesis,2 the possibility of a contribution from stress-related psychosocial factors remains under consideration.3
Over the past 30 years, a body of research has investigated associations between stress-related psychosocial factors and cancer development and progression. Several reviews of this topic have been published, but have produced disparate findings. This outcome might be partly because some reviews have not distinguished between prospective studies and cross-sectional, retrospective casecontrol, or quasi-prospective studies.46 Cross-sectional and retrospective casecontrol studies are subject to recall bias caused by cancer diagnosis or memory distortion, and such studies cannot conclusively detect a longitudinal association between predictors and outcome variables. It is widely recognized in research relating psychological factors to physical illness that patients re-evaluate their lives on the basis of their health state, and might selectively recall their experience before disease diagnosis. For example, if a patient believes that stress causes illness, then he or she may retrospectively reinterpret their earlier experience in order to make sense of the illness. Quasi-prospective studies that have...