We read with interest the skillful meta-analysis by Satin et al 1 indicating that depression is predictive of all-cause mortality, but not disease progression, in cancer patients. We commend the authors for their thoughtful methodology and interpretations, and wish to highlight important questions raised by the article.
How do the effects compare with effects in the general population? To put their findings in context, the authors noted that a comparable meta-analysis found a more robust effect of depression on mortality in patients with cardiac disease. Another meta-analysis (cited by Satin et al but not directly compared with their results) reported an 81% higher mortality rate among subclinically and clinically depressed individuals in community samples, 3 whereas Satin et al found mortality rates to be higher by 39% (for clinical depression) and 25% (for depressive symptoms). 1 Thus, the effect of depression on mortality may not be greater (and potentially somewhat smaller) in cancer patients compared with the general population.
Does depression increase the risk of dying from cancer? Satin et al noted surprise that depression predicted allcause mortality but not disease progression. 1 However, if depression is predictive of all-cause (not cancer-specific) mortality in cancer patients as it is in the general population, this effect may have little to do with the effect of depression on cancer-specific course or outcome (as is aptly pointed out by Satin et al 1 ). As such, the authors' postulate that the nonsignificant link between depression and disease progression in cancer patients is due to the limited number of extant studies should remain speculative until more research has accrued.
What are the implications of the assumed (but controversial) importance of mortality as a primary endpoint of psychosocial oncology interventions? Depression can be lifethreatening, and the findings of Satin et al support the notion that treating depression to prevent mortality is as important for cancer patients as for anyone else. However, the findings do not speak directly to the role of interventions for depression in facilitating survival from cancer.