Depression is a Risk Factor for Mortality in Coronary Heart Disease
Carney, Robert M. PhD; Freedland, Kenneth E. PhD; Sheps, David S. MD, MSPH
Department of Psychiatry; Washington University School of Medicine; St. Louis, MO (Carney, Freedland)
Department of Medicine; Division of Cardiology; University of Florida; Malcom Randall VA Medical Center; Gainesville, FL (Sheps)
Research on the relationship between depression and heart disease has evolved along two parallel paths over the past 15 years. One has examined whether depression is a risk factor for incident coronary heart disease (CHD). Most of these studies have had large, community-based samples, and some have used data from large, multicenter epidemiologic studies such as the landmark Epidemiologic Catchment Area study (1). INTERHEART, one of the largest studies to date to investigate whether depression and other psychosocial factors are associated with incident myocardial infarction (MI), is discussed in a separate editorial in this issue (2–4).
The other path has examined whether depression is a risk factor for cardiovascular morbidity and mortality in the context of established CHD. Most of these studies have been based on much smaller samples drawn from single clinical sites. Some of them have had too few cases and/or too few cardiac events for an adequate test of the hypothesis. For example, one study reported an adjusted odds ratio of 4.9 for mortality in patients with moderate to severe depression, but the result was not statistically significant (5). Although this study had a relatively large sample (560 patients), it had only 12 deaths.
Two independent meta-analyses of studies of depression as a risk factor for incident CHD were published approximately 2 years ago. Rugulies (6) identified 11 studies of depression or depressed mood in initially healthy subjects as predictors of MI or cardiac death. His meta-analysis yielded an aggregate relative risk of 1.64 (95% confidence interval, 1.29–2.08). Wulsin and Singal (7) used slightly different inclusion criteria. They identified 10 studies and also reported an aggregate relative risk of 1.64 (confidence interval, 1.41–1.90).
Two independent meta-analyses are also presented in this issue of Psychosomatic Medicine (8,9). Unlike their predecessors, both examine depression as a predictor of mortality in patients with established CHD. Like their predecessors, the analyses have slightly different inclusion criteria. The criteria for the analysis by Barth et al. (8) were relatively broad in the sense that studies of patients at various stages of CHD were included. They report an odds ratio of 2.24 (1.37–3.60) for mortality in patients with symptoms of depression compared with patients without depression symptoms. Van Melle et al. (9), in contrast, limited their analysis to studies of post-MI patients. They reported odds ratios for all-cause and for cardiac-related mortality of 2.38 (1.76–3.22) and 2.59 (1.77–3.77), respectively. The report by Van Melle et al. (9) is also limited to unadjusted analyses. Barth et al. (8) examined both unadjusted and adjusted analyses but found little difference between them with respect to the risk of mortality.
Having been involved in one of the earliest (10) and in one of the most recent (11) of the studies included in these meta-analyses, we are pleased that both meta-analyses support the conclusion that depression is indeed a risk factor for cardiac and all-cause mortality in patients with established CHD. Rigorous meta-analysis is a difficult and time-consuming endeavor, and we commend the authors of both articles for their efforts. Both groups have done an admirable job of identifying the relevant studies and carefully considering their strengths and limitations when choosing which to include in their analyses.
These meta-analyses are also very timely in light of several recent reports that have failed to find relationships between depression and cardiac mortality in post-MI patients, even in univariate analyses. Mayou et al. (12), for example, reported a nonsignificant 1.6 relative risk of mortality. This study had a small sample size and relied on the Hospital Anxiety and Depression Scale rather than an instrument that has more reliably demonstrated prognostic value in patients with CHD. However, Lane et al. (13) also failed to find a relationship between depression and mortality, despite using the Beck Depression Inventory, a measure that has predicted mortality in a number of other studies.
Barth et al. (8) were unable to find any characteristics differentiating between studies that found or failed to find that depression predicts cardiac events. However, Van Melle et al. (9) noted that the potency of depression as a risk factor may have declined since the mid-1980s, when the earliest studies were conducted. If so, this would be similar to the trend observed more than a decade ago in research on the Type A behavior pattern. Type A was initially identified as a risk factor for incident CHD, but later studies failed to substantiate this (14). Some actually found Type A to be protective in patients with heart disease (15).
One possible explanation for such trends is that publication biases initially favor positive studies; only later, when the research area begins to mature, do failures to replicate earlier, positive findings begin to appear. However, there is another plausible explanation for the declining effect of depression.
In the last 10 to 15 years, there have been many major advances in the treatment of CHD, and the post-MI mortality rate has steadily declined (16). It is possible that the physiological or behavioral pathways that make depression a risk factor for cardiac mortality and morbidity are weakened by treatments that did not exist 10 or 15 years ago but that are now widely administered to post-MI patients. The fact that the effect of depression has apparently declined but not disappeared altogether may favor this explanation.
Another interesting finding concerns the relationship of depression with the risk of mortality over the course of long-term follow-ups. Although Van Melle et al. (9) did not find a difference in odds ratios between follow-ups longer than 6 months and follow-ups of 6 months or less, Barth et al. (8) did find differences among studies that had short-term, medium-term, and long-term follow-ups. The differences depended on how depression was defined. The risk was relatively low in studies of depression symptoms that had long follow-up periods and larger in studies of clinical depression that had long follow-ups. In many cases, depressive symptoms are transient after an acute MI or other cardiac events. Clinical depression, in contrast, tends to be a chronic, recurring condition that could very well continue to affect the course and outcome of CHD long after recovery from an acute MI.
These findings have several implications for future clinical trials. ENRICHD is the only trial published to date that was designed to determine whether treating major or minor depression and/or low perceived social support affects survival after acute MI (17). Treatment for depression was initiated within days or weeks after the acute MI for patients who were randomly assigned to the intervention arm. As is now well known, the ENRICHD intervention had significant effects on depression and social support but not on survival. Given the trends toward declining mortality rates in post-MI patients, and the apparently diminished effects of depression on mortality early after acute MI, future post-MI trials may have to recruit many more patients than previously thought necessary for an adequately powered test of the hypothesis. The patients should probably be recruited several months rather than several days or weeks after the acute event, and they should be followed for at least 2 years after the intervention is terminated. However, the meta-analysis by Barth et al. (8) shows that the prognostic importance of depression is not confined to post-MI patients. Treatment trials are needed for a broader spectrum of cardiac patients, including ones who have been hospitalized for unstable angina, who have undergone revascularization, or who have stable CHD. The fact that mortality rates are even lower in some of these groups than in acute MI patients further underscores the point that large samples will be needed for these trials. Regardless of which group is being studied, the risks associated with clinically significant depression probably cannot be overcome with a short-term intervention. Patients assigned to the intervention arms of future trials should be given not only short-term treatment but also maintenance therapy to prevent depression relapses and recurrences.
Are some measures of depression better predictors than others of the risk for mortality in CHD? Van Melle et al. (9) found no difference in risk whether depression was measured by psychiatric interviews or by self-report inventories. Similarly, Barth et al. (8) found no differences in the level of risk whether depression was defined as a clinical disorder or as a cutoff score on a self-report inventory. Does this mean that planners of future studies should assume that all measures of depression are created equal? Not necessarily. For example, these meta-analyses suggest that the Beck Depression Inventory has a stronger track record of predicting cardiac events than many other self-report depression questionnaires. It would seem prudent to rely whenever possible on structured interviews and self-report inventories that have previously predicted the outcomes under study.
We still do not know which features of depression are associated with an increased risk for mortality and which ones are not. This should be one of the priorities for future research. It is possible that secondary analyses of some of the studies included in these meta-analyses may provide some useful clues. More research is also needed on possible mechanisms underlying the relationship between depression and mortality. If we understood more about how depression confers risk, we would be better able to identify high-risk cases. Furthermore, if successful treatment of depression does not reduce the risk of post-MI mortality, then it may be necessary to redirect our intervention efforts toward the underlying behavioral or pathophysiological mechanisms.
The authors of these new meta-analyses should be commended for their careful work. We believe that they have made important contributions to the field, and that their reports will be cited for many years to come.
Robert M. Carney, PhD
Kenneth E. Freedland, PhD
Department of Psychiatry
Washington University School of Medicine
St. Louis, MO
David S. Sheps, MD, MSPH
Department of Medicine
Division of Cardiology
University of Florida
Malcom Randall VA Medical Center
1. Pratt LA, Ford DE, Crum RM, Armenian HK, Gallo JJ, Eaton WW. Depression, psychotropic medication, and risk of myocardial infarction: prospective data from the Baltimore ECA follow-up. Circulation 1996;94:3123–9.
2. Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lisheng L, INTERHEART Study Investigators. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet 2004;364:937–52.
3. Rosengren A, Hawken S, Ounpuu S, Sliwa K, Zubaid M, Almahmeed WA, Blackett KN, Sitthi-Amorn C, Sato H, Yusuf S, INTERHEART investigators. Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): case-control study. Lancet 2004;364:953–62.
4. Sheps DS, Frasure-Smith N, Freedland KE, Carney RM. The INTERHEART study: intersection between behavioral and general medicine. Psychosom Med 2004;66:797–8.
5. Ladwig KH, Kieser M, Konig J, Breithardt G, Borggrefe M. Affective disorders and survival after acute myocardial infarction. Eur Heart J 1991;12:959–64.
6. Rugulies R. Depression as a predictor for the development of coronary heart disease: a systematic review and meta-analysis of the literature. Am J Prev Med 2003;23:51–61.
7. Wulsin LR, Singal BM. Do depressive symptoms increase the risk for the onset of coronary disease? a systematic quantitative review. Psychosom Med 2003;65:201–10.
8. Barth J, Schumacher M, Herrmann-Lingen C. Depression as a risk factor for mortality in patients with coronary heart disease: a meta-analysis. Psychosom Med 2004;66:802–13.
9. Van Melle JP, de Jonge P, Spiijkerman TA, Tijssen JGP, Ormel J, van Veldhuisen DJ, van den Brink RHS, van den Berg MP. Prognostic association of depression following myocardial infarction with mortality and cardiovascular events: a meta-analysis. Psychosom Med 2004;66:814–22.
10. Carney RM, Rich MW, Freedland KE, teVelde A, Saini J, Simeone C. Major depressive disorder predicts cardiac events in patients with coronary artery disease. Psychosom Med 1988;50:627–33.
11. Carney RM, Blumenthal JA, Catellier D, Freedland KE, Berkman LF, Watkins LL, Czajkowski SM, Hayano J, Jaffe AS. Depression as a risk factor for mortality following acute myocardial infarction. Am J Cardiol 2003;92:1277–81.
12. Mayou RA, Gill D, Thompson DR, Day A, Hicks N, Volmink J, Neil A. Depression and anxiety as predictors of outcome after myocardial infarction. Psychosom Med 2000;62:212–9.
13. Lane D, Carroll D, Ring C, Beevers G, Lip GYH. Mortality and quality of life 12 months after myocardial infraction: effects of depression and anxiety. Psychosom Med 2001;63:221–30.
14. Myrtek M. Meta-analyses of prospective studies on coronary heart disease, type A personality, and hostility. Int J Cardiol 2001;79:245–51.
15. Ragland DR. Type A behavior and outcome of coronary disease. N Engl J Med 1989;321:394.
16. Centers for Disease Control and Prevention. Trends in ischemic heart disease death rates for blacks and whites—United States, 1981–1995. JAMA 1999;281:28–9.
17. The ENRICHD Investigators. Effects of treating depression and low perceived social support on clinical events clinical events after myocardial infarction: the Enhancing Recovery in Coronary Heart Disease (ENRICHD) randomized trial. JAMA 2003;289:3106–16.
Copyright © 2004 by American Psychosomatic Society