Effect of Depression on Five-Year Mortality After an Acute Coronary Syndrome

Transcription

1 Effect of Depression on Five-Year Mortality After an Acute Coronary Syndrome Sherry L. Grace, PhD a,b, *, Susan E. Abbey, MD b,c, Moira K. Kapral, MD, MSc b,c,d, Jiming Fang, PhD d, Robert P. Nolan, PhD, CPsych b,c, and Donna E. Stewart, MD b,c Previous research has established a relation between depression at the time of cardiac hospitalization and patient mortality. The objective of this study was to examine the role of depressive history and symptomatology during hospitalization on 5-year all-cause mortality after admission for an acute coronary syndrome. We recruited 750 patients who had unstable angina pectoris and myocardial infarction from 12 coronary care units between 1997 and Measurements included sociodemographic and clinic data and the Beck Depression Inventory (BDI). Data were linked to an administrative database to determine 5-year all-cause mortality. Survival data were adjusted using a Cox s proportional hazards model. One hundred seventy-four participants (23.2%) self-reported a history of depressed mood for >2 weeks, 235 (31.3%) had elevated BDI scores at index hospitalization, with 105 (14.0%) reporting persistent depressive symptomatology. One hundred fifteen participants (15.3%) died by 5 years after hospitalization. After adjusting for prognostic indicators, such as cardiac disease severity, medical history, and smoking, depressive symptomatology during hospitalization was significantly predictive of mortality, but depressive history was not. Hazard ratios associated with BDI scores <10 versus those >10 at hospitalization ranged from 1.90 (95% confidence interval 1.12 to 3.24) at 2 years to 1.53 (95% confidence interval 1.04 to 2.24) at 5 years. In conclusion, the significance of depressive symptomatology at the time of, but not before, hospitalization underlines the need for early identification of increased distress and renews calls to identify treatments that not only improve quality of life but also decrease the risk of mortality Elsevier Inc. All rights reserved. (Am J Cardiol 2005;96: ) It is depressive symptomatology as assessed during hospitalization that is so robustly related to long-term mortality after an acute coronary syndrome (ACS). 1 Most studies that have demonstrated the relation between depression and mortality have assessed depression only 1 time. However, it is well documented that a history of depression is related to future depressive episodes 2 (even among patients who have cardiac problems 1,3 6 ) and to the onset of ACS. 7 9 This study examined the concurrent effects of depressed mood before ACS and depressive symptomatology at the index cardiac event on all-cause mortality across a 5-year period after the index hospitalization. We postulated that such Methods Design and procedure: This was a longitudinal observational study that followed participants from admission for myocardial infarction or unstable angina for 5 years. Details concerning the methods of this study have been described previously. 10 Eligible patients were recruited between August 1997 and January 1999 in the coronary care unit by research nurses on the second to fifth days of hospitalization. Those who met the study criteria and agreed to participate signed a consent form and were provided with a self-report questionnaire. Consent was also obtained to link participants self-report questionnaire data with their clinical data. Data were later linked to population-based administrative databases using unique encrypted identifiers, including the Ontario Registered Persons Database and the Discharge Abstract Database of the Canadian Institute of Health Information. Measurements: The self-report survey completed in the hospital consisted of questions about sociodemographic characteristics, medical history, and depressive symptoma York University, b University Health Network; c University of Toronto; and d Institute for Clinical and Evaluative Services, Toronto, Ontario, Canada. Manuscript received April 13, 2005; revised manuscript received and accepted June 20, Drs. Abbey and Stewart were supported by research grants from The Heart and Stroke Foundation of Ontario, Ottawa, Ontario, and Samuel Lunenfeld Foundation, Toronto, Ontario, Canada. Dr. Kapral is supported by the University Health Network Women s Health Program and the Canadian Stroke Network, Toronto, Ontario, Canada. * Corresponding author: Tel: , ext 22364; fax: address: sgrace@yorku.ca (S.L. Grace). persistent depressive symptomatology would have detrimental effects on mortality /05/$ see front matter 2005 Elsevier Inc. All rights reserved. doi: /j.amjcard

2 1180 The American Journal of Cardiology ( Table 1 Nurse-reported characteristics of participants versus nonparticipants (ineligible, refused) at time of index hospitalization Variable Participants (n 910) Nonparticipants (n 890) Women 320 (35.0%) 385 (43.7%)* Married 667 (73.1%) 551 (62.5%)* Myocardial infarct 478 (52.4%) 490 (55.6%) Age (yrs) 62 (12.00) 69 (12.25)* *p atology. The sociodemographic data included age, gender, marital status, education level, and family income. Participants were asked whether they had a history of depressed mood for 2 weeks (yes/no). Survey data were linked with clinical data that were compiled by coronary care unit nurses, which included Killip s class 11 as an indicator of illness severity and medical history (i.e., hypertension, diabetes, congestive heart failure, or previous myocardial infarction). Depressive symptoms at the time of hospitalization were assessed by the Beck Depression Inventory (BDI), 12 a reliable and well-validated 21-item scale that uses a forcedchoice 4-alternative response format that has been widely used in the general population and in populations with long-term illness, including cardiac problems Higher scores reflect greater depressive symptomatology, with scores 10 reflecting mild to severe symptomatology. The internal consistency of the BDI had an value of 0.87 in the present sample. Based on the universal and centralized health care system in the province, we were able to obtain comprehensive follow-up information on the sample of participants using the Registered Persons Database. Linkage to a Discharge Abstract Database enabled determination of co-morbidity to compute the Charlson Comorbidity Index. 18 Participants: Consecutive patients who were diagnosed with myocardial infarction or unstable angina in 12 coronary care units across south-central Ontario, Canada, were approached for the study. Characteristics of participants and nonparticipants are listed in Table 1. Of the almost 1,800 patients who were approached, participants were significantly younger than those who refused or were ineligible to participate (t 13.08, p 0.001). Significantly more men than women agreed to participate (chi-square 14.00, p 0.001), and more married and fewer widowed patients agreed to participate (chi-square 21.46, p 0.001). Nine hundred ten patients (590 men and 320 women, 69% response rate) consented to participate in the study. Patients were 31 to 93 years old (mean ). Seventy-four percent of participants (n 664) were married, and 45% (n 340) had a family income above $50,000 Canadian (i.e., $32,000 United States) annually. Fifty-three percent (n 482) had a confirmed myocardial infarction, and 47% (n 424) were diagnosed as having unstable angina after appropriate diagnostic tests. From the initial cohort of 910 participants who were recruited, we were able to match 867 of these participants with unique encrypted identifiers (43 participants lost, 4.7%), and link 856 to the Canadian Institute of Health Information Discharge Abstract Database (11 patients lost, 1.2% of overall cohort). No cases were lost while linking to the Registered Persons Database. Of this cohort, 836 had complete baseline BDI depression scores (20 patients lost, 2.2% of overall cohort), and 750 had a valid history of depression responses (86 patients lost, 9.5% of overall cohort). Statistical analyses: Statistical analyses were performed with SAS 8.2 (SAS Institute, Cary, North Carolina). Data were cleaned and screened to evaluate statistical assumptions. A descriptive examination of the variables was performed, and baseline participant characteristics were compared based on differences in mortality and depressive symptomatology (BDI score 10 vs 10). For survival analysis, time to event was measured from the index hospitalization. Cox s proportional hazard model was used to estimate the hazard ratio and 95% confidence intervals for all-cause mortality as based on history of depressed mood, depressive symptomatology at index hospitalization, and their interaction. A backward stepwise model was computed, with 0.1 as the significance level for keeping an explanatory variable in the model. Within Cox s model, we compared all-cause mortality rates with adjustment for the prognostic factors: gender, age, education level, family income, marital status, smoking status, Killip s class, diagnosis (myocardial infarction or unstable angina), hypertension (systolic blood pressure 130 mm Hg), previous myocardial infarction or congestive heart failure, diabetes, and the Charlson Co-morbidity Index (0 vs 1 co-morbid condition). Results Differences in participant characteristics based on self-reported depressive symptomatology at index hospitalization: As presented in Table 2, participants who had a BDI score 10 during hospitalization were more likely to be women, have lower family income, be widowed, separated, or divorced, have fewer years of education, be a smoker, have a diagnosis of unstable angina, have greater disease severity (as operationalized by Killip s class), have more co-morbid conditions, and have a medical history of diabetes and congestive heart failure. Relation of timing and persistence of depressive symptomatology to all-cause mortality: Depressive symptomatology before and at the time of the index ACS event is presented in Table 3. During the 5-year follow-up, there were 115 deaths (Table 2). Analyses showed an interaction between depressive history and depressive symptomatology during hospitalization on 5-year mortality. As presented in Table 3, the lowest mortality rate was found in participants who had a history of depressive symptomatology and no

3 Coronary Artery Disease/Depression and Mortality in Cardiac Patients 1181 Table 2 Participant characteristics of those with subthreshold depressive symptomatology (Beck Depression Inventory 10) versus those with depressive increased symptomatology (Beck Depression Inventory 10) at index hospitalization Overall Depressive Symptoms p Value Subthreshold (BDI 10) High (BDI 10) Sample (68.7%) 235 (31.3%) Men 486/750 (64.8%) 376/515 (73%) 110/235 (46.8%) Age (yrs) Killip s class I 591/711 (83.1%) 414/486 (85.2%) 177/225 (78.7%) Family income $50,000 Canadian 359/646 (55.6%) 231/451 (51.2%) 128/195 (65.6%) Married 555/744 (74.6%) 399/510 (78.2%) 156/234 (66.7%) Education level below high school 506/738 (68.6%) 336/510 (65.9%) 170/228 (74.6%) Index diagnosis of myocardial infarction 392/750 (52.3%) 294/515 (57.1%) 98/235 (41.7%) Systolic blood pressure 130 mm Hg 309/719 (43%) 214/490 (43.7%) 95/229 (41.5%) Charlson Co-morbidity Index 1 531/750 (70.8%) 384/515 (74.6%) 147/235 (62.6%) Diabetes mellitus 182/750 (24.3%) 108/515 (21%) 74/235 (31.5%) Previous myocardial infarction 537/750 (71.6%) 378/515 (73.4%) 159/235 (67.7%) Congestive heart failure 88/750 (11.7%) 47/515 (9.1%) 41/235 (17.4%) Smoker 241/745 (32.3%) 152/514 (29.6%) 89/231 (38.5%) Length of hospital stay BDI History of depressed mood 174/750 (23.2%) 69/515 (13.4%) 105/235 (44.7%) Death status Died 1 yr after discharge 39/750 (5.2%) 20/515 (3.9%) 19/235 (8.1%) Died 2 yrs after discharge 56/750 (7.5%) 30/515 (5.8%) 26/235 (11.1%) Died 3 yrs after discharge 77/750 (10.3%) 45/515 (8.7%) 32/235 (13.6%) Died 4 yrs after discharge 95/750 (12.7%) 57/515 (11.1%) 38/235 (16.2%) Died 5 yrs after discharge 115/750 (15.3%) 69/515 (13.4%) 46/235 (19.6%) Values are means SD or numbers of patients (percentages). increased depressive symptoms at the time of index hospitalization, whereas the highest mortality rate occurred in participants who had no history of depressed mood and developed increased depressive symptoms at the time of the index hospitalization. Based on pairwise comparisons (Table 4), those who had increased depressive onset only at the time of the index hospitalization were at significantly greater risk of mortality than were participants who had no depressive symptomatology at any time point or those who previously had only a depressed mood. With regard to depressed mood, participants who had depressive symptoms onset at the index hospitalization were 4 times more likely to die (hazard ratio 4.6). The stepwise Cox s proportional hazards models for allcause mortality across 5 years after index hospitalization are presented in Table 5, after adjusting for gender, age, family income, marital status, education level, diagnosis, high systolic blood pressure, smoking, Killip s class, co-morbid conditions, and a medical history of diabetes, myocardial infarction, or congestive heart failure. Depressive symptomatology at the time of index hospitalization remained a significant predictor of all-cause mortality across years 2 through 5; however, depressive history and its interaction with depressive symptomatology at the time of hospitalization did not remain in the adjusted models during any year of follow-up. The hazard ratio associated with subthreshold symptomatology (BDI score 10) compared with increased depressive symptomatology (BDI score 10) at the time of hospitalization ranged from 1.90 at 2 years to 1.53 at 5 years after index hospitalization. Kaplan-Meier adjusted survival curves for 5-year all-cause mortality by age, systolic blood pressure, co-morbidity, and medical history of congestive heart failure using the corrected group prognosis method are presented in Figure 1. Discussion Depressive symptomatology and major depressive disorders are common among patients who have ACS. 10,19 Similar to previous findings, 5 23% of participants reported a history of depressed mood. The prevalence of major depression at the time of hospitalization ranges from 15% to 20% in the literature, 20 which is approximately threefold higher than those of age-matched community-based prevalence studies. 21,22 In our sample, 31% had increased depressive symptomatology at the time of hospitalization, 14% of whom had a persistent depressed mood before the index hospitalization. We were interested in the prognostic significance of historical and persistent depressive symptomatology on mortality to build on the findings of Lesperance et al. 1,5 Although a strong body of literature demonstrates that depressive history is related to onset of ACS, 7 9 our results suggest that this history is not as salient to mortality among patients who have established disease compared with de-

4 1182 The American Journal of Cardiology ( Table 3 Participants who were alive versus those who died five years after discharge Variable Overall Status p Value Alive Died Sample (84.7%) 115 (15.3%) Men 486/750 (64.8%) 418/486 (86%) 68/486 (14%) Women 264/750 (35.2%) 217/264 (82.2%) 47/264 (17.8%) Age (yrs) Killip s class I 591/711 (83.1%) 515/591 (87.1%) 76/591 (12.9%) I 120/711 (16.9%) 86/120 (71.7%) 34/120 (28.3%) Family income $50, /646 (55.6%) 298/359 (83%) 61/359 (17%) $50, /646 (44.4%) 255/287 (88.9%) 32/287 (11.1%) Marital status Not married 189/744 (25.4%) 145/189 (76.7%) 44/189 (23.3%) Married 555/744 (74.6%) 485/555 (87.4%) 70/555 (12.6%) Education level Below high school 506/738 (68.6%) 418/506 (82.6%) 88/506 (17.4%) Above high school 232/738 (31.4%) 206/232 (88.8%) 26/232 (11.2%) Index diagnosis Unstable Angina 392/750 (52.3%) 331/392 (84.4%) 61/392 (15.6%) Myocardial Infarction 358/750 (47.7%) 304/358 (84.9%) 54/358 (15.1%) Systolic blood pressure 130 mm Hg 309/719 (43%) 250/309 (80.9%) 59/309 (19.1%) 130 mm Hg 410/719 (57%) 357/410 (87.1%) 53/410 (12.9%) Charlson Co-morbidity Index 1 531/750 (70.8%) 484/531 (91.1%) 47/531 (8.9%) 1 219/750 (29.2%) 151/219 (68.9%) 68/219 (31.1%) Diabetes mellitus 182/750 (24.3%) 143/182 (78.6%) 39/182 (21.4%) Previous myocardial infarction 537/750 (71.6%) 445/537 (82.9%) 92/537 (17.1%) Congestive heart failure 88/750 (11.7%) 45/88 (51.1%) 43/88 (48.9%) Smoker 241/745 (32.3%) 216/241 (89.6%) 25/241 (10.4%) Length of hospital stay BDI History of depressed mood 174/750 (23.2%) 155/174 (89.1%) 19/174 (10.9%) BDI score /750 (68.7%) 446/515 (86.6%) 69/515 (13.4%) /750 (31.3%) 189/235 (80.4%) 46/235 (19.6%) History of depressed mood BDI* 1(0 0) 446/750 (59.5%) 381/446 (85.4%) 65/446 (14.6%) (0 1) 130/750 (17.3%) 99/130 (76.2%) 31/130 (23.8%) 3(1 0) 69/750 (9.2%) 65/69 (94.2%) 4/69 (5.8%) 4(1 1) 105/750 (14%) 90/105 (85.7%) 15/105 (14.3%) Values are means SD or numbers of patients (percentages). *1 no history of depression and BDI score 10; 2 no history of depression and BDI score 10; 3 history of depression and BDI score 10; 4 history of depression and BDI score 10. Table 4 Pairwise comparisons of interaction between history of depression and presence of increased depressive symptomatology at index hospitalization Pairwise Comparison* Chi-square p Value Hazard Ratio 95% Confidence Interval (0,1) vs (0,0) (1,0) vs (0,0) (1,1) vs (0,0) (0,1) vs (1,0) (0,1) vs (1,1) (1,0) vs (1,1) *0 depression absent; 1 depression present. pressive severity at the time of hospitalization. This was also shown in a study by Berkman et al, 23 where depressive symptomatology was assessed up to 3 years before a myocardial infarction and was not related to mortality 6 months after a myocardial infarction. In addition, Bush et al 3 found no difference in mortality risk at 4 months based on a history of depression. With regard to the persistence of depressive symptomatology, the prognostic implications of this have been relatively unexplored. 24 Contrary to expectation, our results reveal that persistent depressive symptomatology that recurs at the time of hospitalization is not of prognostic importance

5 Coronary Artery Disease/Depression and Mortality in Cardiac Patients 1183 Table 5 Adjusted stepwise Cox s model by year since index hospitalization Years Since Index Hospitalization Variables Chi-Square p Value Hazard Ratio 95% Confidence Intervals 1 BDI score Age Systolic blood pressure 130 mm Hg Charlson Co-morbidity Index BDI score Age Congestive heart failure BDI score Age Systolic blood pressure 130 mm Hg Congestive heart failure BDI score Gender Age Marital status Systolic blood pressure 130 mm Hg Congestive heart failure BDI score Age Systolic blood pressure 130 mm Hg Charlson Co-morbidity Index Congestive heart failure Figure 1. Kaplan-Meier adjusted survival curves for 5-year all-cause mortality by a history of depression (History Dep) and depressive symptomatology at the time of index hospitalization (In Hosp; BDI score 10 vs 10). compared with severity only at the time of hospitalization. This runs counter to findings from the cohort presented by We also created a variable for no increased depressive symptomatology at any time point, depressive symptomatology before or at index hospitalization, and depressive symptomatology at the 2 times. This was not related to mortality in the Cox s models. Lesperance et al, 5 in which patients who had recurrent depressive symptomatology at the time of hospitalization were at significantly increased risk of 18-month mortality compared with those who had depressive onset only at the time of hospitalization. However, their analysis was not conducted within an adjusted Cox s model. Further research using longitudinal community cohorts and diagnostic as-

6 1184 The American Journal of Cardiology ( sessments is needed to elucidate the role of depression chronicity and timing in mortality. The significant interaction between history of depressed mood and depressive symptomatology at the time of hospitalization is intriguing and replicates the findings of Lesperance et al. 5 They also found that the group with the lowest risk of mortality consisted of patients who had a history of depression and did not report increased depressive symptomatology in the hospital. Lesperance et al speculated that those who had a history of depression and managed to resist the stresses of ACS had enhanced physical or psychic resources that enabled them to survive. However, in the full Cox s model, this interaction was superceded by the importance of depressive symptomatology at the time of hospitalization. Nevertheless, this interaction deserves further thought and inquiry, particularly with regard to mechanisms. The negative prognostic effect of acute onset depressive symptomatology at the time of hospitalization raises further questions with regard to pathophysiologic, 20 rather than behavioral, mechanisms. These results corroborate that even mild depressive symptoms at the time of cardiac hospitalization are not benign symptoms solely attributable to a life-threatening event, somatic complaints, or hospital setting but are robust symptoms with important prognostic implications for years to follow. They also highlight the importance of early identification of depressive symptomatology during hospitalization. Once identified, patients who have increased symptomatology should be offered evidence-based treatments. 25 Although treating depression in the context of ACS does improve quality of life and functional capacity, 25 it is unclear whether treating depression may have beneficial effects on cardiac prognosis. 20,26 There is much current research that is examining potential behavioral and physiologic mechanisms that link depression to prognosis and how psychotherapeutic and psychopharmacologic treatment might mitigate this relation. Caution is warranted when interpreting these results. First, the generalizability of the study is limited by the differences between participants and refusers and the loss of participants in the linkage process. However, we adjusted for these differences in the Cox s model to statistically control for these biases. Second, our self-report measurement of depressive symptomatology assessed severity only and does not provide diagnostic information. Moreover, our self-report measurement of depressive history may be hampered by retrospective bias and was not verified through a structured diagnostic interview. However, 2 recent metaanalyses concluded that assessments by self-report questionnaires of symptomatology and clinical interviews based on diagnostic criteria are related to mortality and that, for these purposes, neither type of depressive assessment is superior. 27,28 Our outcome of mortality was not broken down by cardiac and noncardiac causes but was based on all-cause mortality only. Third, our results are limited by the small number of deaths in the group of participants who had a history of depressed mood and BDI scores 10. Further examination and replication in a larger population-based sample is warranted. Acknowledgment: We acknowledge Linda Green, PhD, and the coronary care unit nurses for their diligence in participant recruitment and Claus Wall, MSc, for data linkage. 1. Lesperance J, Frasure-Smith N, Talajic M, Bourassa M. Five-year risk of cardiac mortality in relation to initial severity and one-year changes in depression symptoms after myocardial infarction. Circulation 2002; 105: Keller M, Lavori P, Lewis CE, Klerman GL. Predictors of relapse in major depressive disorder. JAMA 1983;250: Bush D, Ziegelstein R, Tayback M, Richter D, Stevens S, Zahalsky H, Fauerbach JA. Even minimal symptoms of depression increase mortality risk after myocardial infarction. Am J Cardiol 2001;88: Hance M, Carney RM, Freedland KE, Skala J. Depression in heart patients with coronary heart disease: a 12-month follow-up. Gen Hosp Psychiatry 1996;18: Lesperance F, Frasure-Smith N, Talajic M. Major depression before and after myocardial infarction: its nature and consequences. Psychosom Med 1996;58: Schrader G, Cheok F, Hordacre A-L, Guiver N. Predictors of depression three months after cardiac hospitalization. Psychosom Med 2004; 66: Hemingway H, Marmot M. Evidence based cardiology: psychosocial factors in the aetiology and prognosis of coronary heart disease: systematic review of prospective cohort studies. BMJ 1999;318: Rugulies R. Depression as a predictor for coronary heart disease: a review and meta-analysis. Am J Prev Med 2002;23: Wulsin L, Singal B. Do depressive symptoms increase the risk for the onset of coronary disease? A systematic quantitative review. Psychosom Med 2003;65: Grace SL, Abbey S, Pinto R, Shnek Z, Irvine J, Stewart DE. Longitudinal course of depressive symptomatology following a cardiac event: effect of gender and cardiac rehabilitation. Psychosom Med 2005;67: Killip T, Kimball JT. Treatment of myocardial infarction in a coronary care unit: a two year experience with 250 patients. Am J Cardiol 1967;20: Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961;4: Steer RA, Cavalieri TA, Leonard DM, Beck AT. Use of the Beck Depression Inventory for primary care to screen for major depression disorders. Gen Hosp Psychiatry 1999;21: Frasure-Smith N, Lesperance F, Talajic M. Depression following myocardial infarction: impact on six-month survival. JAMA 1993;270: Shnek ZM, Irvine J, Stewart DE, Abbey S. Psychological factors and depressive symptoms in ischemic heart disease. Health Psychol 2001; 20: Beck AT, Steer RA, Garbin MG. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clin Psychol Rev 1988;8: Frasure-Smith N. In-hospital symptoms of psychological stress as predictors of long-term outcome after acute myocardial infarction in men. Am J Cardiol 1991;67: Charlson M, Pompei P, Ales K, MacKenzie C. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:

7 Coronary Artery Disease/Depression and Mortality in Cardiac Patients Abbey SE, Stewart DE. Gender and psychosomatic aspects of ischemic heart disease. J Psychosom Res 2000;48: Lett H, Blumenthal J, Babyak M, Sherwood A, Strauman T, Robins C, Newman NF. Depression as a risk factor for coronary artery disease: evidence, mechanisms, treatment. Psychosom Med 2004;66: Blazer DG, Kessler RC, McGonagle K, Swartz M. The prevalence and distribution of major depression in a national community sample: the national comorbidity survey. Am J Psychiatry 1994;151: Lavie C, Milani R, Cassidy M, Gilliland Y. Effects of cardiac rehabilitation and exercise training programs in women with depression. Am J Cardiol 1999;83: Berkman LF, Leo-Summers L, Horwitz RI. Emotional support and survival after myocardial infarction: a prospective, population-based study of the elderly. Ann Intern Med 1992;117: Burg M, Benedetto M, Soufer R. Depressive symptoms and mortality two years after coronary artery bypass surgery in men. Psychosom Med 2003;65: Swenson J, O Connor S, Barton D, Van Zyl L, Swedberg K, Forman LM, Gaffney M, Glassman AH. Influence of depression and effect of treatment with sertraline on quality of life after hospitalization for acute coronary syndrome. Am J Cardiol 2003; 92: Davidson K, Rieckmann N, Lesperance J. Psychological theories of depression: potential application for the prevention of acute coronary syndrome recurrence. Psychosom Med 2004;66: Barth J, Schumacher M, Herrmann-Lingen C. Depression as a risk factor for mortality in patients with coronary heart disease: a metaanalysis. Psychosom Med 2004;66: van Melle J, De Jonge P, Spijkerman T, Tjisssen J, Ormel J, van Veldhuisen D, 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: