This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Coryell, W. Articles by Keller, M. Search for Related Content PubMed PubMed Citation Articles by Coryell, W. Articles by Keller, M. Related Collections Depression Other Cardiovascular Medicine

Persistence of Depressive Symptoms and Cardiovascular Death Among Patients With Affective Disorder

From the Department of Psychiatry (W.C., C.T.), University of Iowa, Iowa City, IA; Department of Psychiatry (A.L.), Cornell Medical College, Ithaca, NY; National Institute of Mental Health (J.D.M.), Bethesda, MD; Rhode Island Hospital (D.S.), Providence, RI; Department of Research Assessment and Training (J.E.), New York State Psychiatric Institute, New York; and Butler Hospital (T.M., M.K.), Providence, RI.

Address reprint requests to: William Coryell, MD, University of Iowa College of Medicine, Psychiatry Research–MEB, Iowa City, IA 52242-1000. Email: william-coryell{at}uiowa.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
OBJECTIVE: Studies of both community and clinical samples have associated depressive symptoms with risks for subsequent cardiovascular morbidity and mortality. Because the physiological mechanisms thought to underlie this link would be cumulative in their effects, the following analyses tested the prediction that risks for cardiovascular death would increase in proportion to the persistence of depressive symptoms in a long-term follow-up.

METHODS: Baseline assessment was performed as patients sought treatment for major depressive disorder, mania, or schizoaffective disorder. Follow-up evaluations occurred semiannually for the next 5 years and annually thereafter. The 903 patients described, observed for a mean of 11.0 years (SD = 5.2 years), were divided into thirds according to the proportion of follow-up weeks in episodes of major depressive disorder, schizoaffective disorder, or intermittent depressive disorder. The resulting groups were then compared by cumulative risks of cardiovascular death.

RESULTS: Patients whose depressive symptoms were the most persistent were no more likely to die of cardiovascular causes than were those with the fewest weeks ill. A regression analysis showed that older age and the presence of cardiovascular disease at baseline, but not the subsequent chronicity of depressive symptoms, predicted cardiovascular death.

CONCLUSIONS: The physiological concomitants of depressive illness apparently do not promote cardiovascular mortality in a cumulative manner. Efforts should be directed toward identification of risk factors common to both lifetime depressive symptoms and cardiovascular morbidity.

Key Words: affective disorder • cardiovascular mortality • follow-up

Abbreviations: MDD = major depressive disorder; RDC = ResearchDiagnostic Criteria.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Studies have linked affective illness and cardiovascular mortality in a number of ways. Within community samples, individuals with higher scores on scales measuring depressive symptoms in general (1, 2), or hopelessness in particular (3), were shown to be more likely to die during subsequent follow-up periods. The diagnosis of depressive disorder has also predicted excess cardiovascular mortality in other community samples (4, 5). Congruent findings have emerged from clinical samples of depressed patients as well (6, 7), although some studies have found no excess in cardiovascular mortality (8–10) and at least one noted a significant deficit in such deaths among depressed patients (11). Other work has linked depressive symptoms to earlier death among patients who recently had a myocardial infarction (12, 13), patients with clinically diagnosed hypertension (14), and patients with coronary artery disease (15).

An apparent association between depressive and cardiovascular morbidity may arise in various ways. The physical limitations resulting from cardiovascular illness are demoralizing and may in this way give rise to depressive symptoms. Cerebrovascular changes may also produce depressive illness more directly (16, 17). Moreover, depressive disorder and cardiovascular illness may share important risk factors, such as smoking, diabetes, and obesity. Notably, though, several efforts to statistically control for baseline risk factors have not eliminated, or even substantially lessened, the association between cardiovascular morbidity and depressive symptoms (2–4). Another study found that controlling for alcoholism and smoking eliminated the apparent contribution of affective disorder to mortality from natural causes (18). The sample contained only 25 individuals with affective spectrum disorders, though, and statistical power was therefore quite limited.

Another possible factor, particularly relevant in clinical samples, is the potential cardiotoxicity of tricyclic antidepressants, the medications most often used for depressive disorders before the advent of serotonin reuptake inhibitors. Of the clinical studies listed above, though, only one directly addressed this issue. Weeke et al. (7) subdivided their cohort into tricyclic era and pre–tricyclic era subgroups. An excess in cardiovascular mortality was seen in both groups and was slightly, but not significantly, greater among the pre–tricyclic era patients.

The possibility that the physiological changes attending affective disorder may promote cardiovascular morbidity has received particular attention. Both depression and mania are associated with increased sympathetic activity and this, in turn, predisposes to ventricular arrhythmias and low heart rate variability (15, 19). The stress resulting from depression and/or the serotonin abnormalities underlying it may also increase platelet coagulability (20, 21). In either case, mechanisms by which depression would lead to cardiovascular morbidity would be cumulative. If depressive illness predisposes to ventricular arrhythmia, symptom persistence would increase cardiovascular mortality by extending periods at risk. The effects of depression on platelet coagulability would operate in a more insidious manner by gradually hastening the formation of arterial plaque. Either mechanism raises the possibility that effective treatment of affective disorders may not only result in relief from the well-recognized impairments arising directly from the syndrome but may, in addition, lessen eventual risks for cardiovascular illness or complications thereof.

Most studies have identified depressive symptoms at one point and cardiovascular outcomes at another. If affective illness per se promotes cardiovascular morbidity, and if the successful treatment of depression would therefore lessen it, then some relationship between the persistence of depressive syndromes and cardiovascular morbidity should be demonstrable.

The following analyses make use of data from a long-term, high-surveillance follow-up of patients with major affective disorder. Because the cohort spanned a wide age range and has now been followed for 15 years and beyond, a number of deaths, many of them due to cardiovascular causes, have accumulated. If depressive symptoms promote cardiovascular morbidity in a cumulative manner, then there should be a relationship between the prospectively observed persistence of depressive symptoms and the risk of death due to cardiovascular causes.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Subjects
The National Institute of Mental Health Collaborative Program on the Psychobiology of Depression –Clinical Studies identified patients who met RDC (22) for MDD, mania, or schizoaffective disorder as they sought treatment at any of five academic treatment centers: Harvard University, Boston, MA; Rush-Presbyterian St. Luke’s Medical Center, Chicago, IL; University of Iowa College of Medicine, Iowa City, IA; Columbia University, New York, NY; and Washington University School of Medicine, St. Louis, MO. Entry criteria required that participants be white, 18 years of age or older, English-speaking, and knowledgeable of their biological parents. Medical illnesses were not grounds for exclusion, nor was willingness to comply with treatment a factor in participation. The following analyses include all 903 subjects who had at least 6 months of follow-up. Thus, the sample is diagnostically diverse, but all individuals had in common the presence, at entry, of a major affective disorder.

Procedures
All subjects provided written informed consent before baseline assessment began. Information from medical records, interviews with informants, and a structured interview of the patient with the full Schedule for Affective Disorders and Schizophrenia (23) were used to determine the psychiatric diagnosis. The baseline assessment battery included the Admission Medical History Form (available on request), a series of questions concerning the lifetime presence or absence of various medical illness, and the use of various nonpsychotropic medications in the month preceding entry.

Raters contacted patients at 6-month intervals for the first 5 years of follow-up and annually thereafter. They used the Longitudinal Interval Follow-up Evaluation (24) to track each RDC syndrome and based ratings on both the patient interview and a review of medical records. Each of the major RDC syndromes, MDD, manic disorder, and schizoaffective disorder, depressed or manic, were assigned psychiatric status ratings ranging from 1 to 6. A score of 3 or more indicated that the syndrome remained active, and scores of 5 or 6 implied full syndromes at the definite level. Recovery required ratings of 1 or 2 for 8 consecutive weeks, and subsequent ratings above 2 required reappearance of the full syndrome. Intermittent depressive disorder and minor depressive disorder were rated on three-point scales, as were such nonaffective illnesses as alcoholism and anxiety disorders.

When efforts to locate a subject revealed that he or she had died, the rater gathered pertinent medical records, attempted to interview family members about the circumstances of death, and, in many cases, obtained death certificates. Raters, though, did not attempt to fill in symptom ratings for the weeks since the last personal interview. Thus, symptom ratings were absent for periods of up to 52 weeks preceding death. Information from all sources were integrated to complete the suicidal report or cause of death report (GRIM; available on request). In addition to death by suicide and accident, this form listed 19 other specific causes of death. There was only one category for cardiovascular death.

Although study participation did not influence treatment, raters carefully monitored the types and quantities of somatic treatments received during the follow-up intervals. The daily dose for each psychoactive drug was recorded week by week, and the resulting data were later reduced to drug class equivalencies. These were finally converted to composite antidepressant scores as described elsewhere (25).

Analysis
The following analyses used morbidity indexes to quantify the persistence of symptoms during follow-up. For most analyses, the morbidity index was the proportion of follow-up weeks during which psychiatric status ratings for MDD, schizoaffective depression, intermittent depressive disorder, or minor depressive disorder exceeded 2 (ie, the proportion of time during which even mild depressive syndromes were present). Other morbidity indexes were calculated, as described below, to also capture symptom severity and to include weeks during which manic or hypomanic syndromes were present.

Morbidity indexes reflected symptoms for as long as patients remained under observation. To determine the effects of symptom persistence, patients were grouped by morbidity index values into low, medium, and high thirds. Life table methods (26, 27) were then used to compare these three groups by times to cardiovascular death. Kaplan-Meier estimates therefore incorporated censored values and reflected all longitudinal data. Because the observation periods varied among patients, the proportions reported as dying of cardiovascular causes are Kaplan-Meier product limit estimates. Survival curves were compared using log-rank tests (28).

The potential effects of antidepressant treatment were considered in two ways. First, the proportion of weeks during which patients took any drug commonly used in the management of depression was determined. These drugs are included in the composite antidepressant score and are listed elsewhere (25). Patients were grouped by thirds according to this measure, and groups were then compared by estimated times to cardiovascular death, as was done with morbidity index thirds. Second, to quantify the likelihood of antidepressant treatment given the presence of depressive symptoms, the ratio of weeks ill with antidepressant treatment to total weeks ill was determined for each individual, and, as before, thirds derived from these values were compared by time to cardiovascular death.

² tests and t tests were used to compare the demographic and clinical characteristics of those who did and did not die of cardiovascular causes. Cox proportional hazards models were used to examine the strength of the association between hypothesized risk factors and cardiovascular death over the follow-up period. A two-tailed level of 0.05 was used for each statistical test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Table 1 describes the 903 patients who had at least 6 months of follow-up; 529 (58.6%) of these were women and 374 (41.4%) were men. During a mean observation period of 11.0 years (SD = 5.2 years), 51 patients died of cardiovascular causes. In comparison with all other patients, including those who died of other causes, they were older at entry and were more likely to have reported cardiovascular disease. They were also older at the first episode of major affective disorder, more likely to be widowed, and less likely to be single. Patients from New York were overrepresented, and those from Iowa underrepresented, among those who died of cardiovascular causes. Those who died were not distinguished by sex, by the proportions with bipolar I or bipolar II disorders, or by the proportions of alcoholism or drug abuse active at the beginning of follow-up.

View larger version (21K): [in this window] [in a new window]   Fig. 1. Cumulative probabilities of death due to cardiovascular causes by depressive morbidity index group.

The contributions of potential risk factors for cardiovascular death were quantified by logistic regression (Table 2). With age at entry, cardiovascular disease reported at entry, sex, and morbidity index group in the model, only the first two variables were significant risk factors, and there were no statistically significant interactions between risk factors. The addition of active alcoholism and, separately, active drug abuse, did not alter the risks associated with age or reported cardiovascular illness at entry (not shown). Nor did the addition of marital status (single vs. others and widowed vs. others), center (New York vs. others and Iowa vs. others), or age at onset of first major affective disorder episode alter these findings. In each of these repeated analyses, age at entry and cardiovascular disease at entry remained the only significant predictors, and their robustness did not decrease substantially.

We also considered whether depressive morbidity was related to death from natural causes generally. Estimated risks (SE) at 10 years for deaths due to all causes other than accident or suicide were 7.5% (1.6), 6.0% (1.5), and 9.3% (2.1), respectively. The differences were not significant.

Patients most likely to have been in treatment in any given week of follow-up, those in the highest treatment third, were no more or less likely to die of cardiovascular causes than were patients in the lowest treatment third. Men and women in the former group received some antidepressant therapy in 89.5% (SD = 10.9%) and 87.0% (SD = 11.2%) of the weeks of follow-up, respectively. Men and women in the lowest third received treatment during 2.9% (SD = 2.9%) and 4.5% (SD = 4.4%) of weeks, respectively. Likewise, the ratio of weeks in which depressive symptoms and antidepressant treatment coexisted to total weeks with depressive symptoms bore no relationship to risk of cardiovascular death in either sex.

To determine whether feelings of hopelessness per se, as reported at entry, were predictive of subsequent cardiovascular deaths, patients were grouped according to their ratings on the Schedule for Affective Disorders and Schizophrenia hopelessness item (item 244) into those with none or slight discouragement about the future, those with "mild" or "moderate" pessimism, and those with "severe" or "extreme" feelings of pessimism. Patients who began the study with low (N = 137), medium (N = 352), or high (N = 414) levels of pessimism did not differ by risk of cardiovascular death. The estimated hazards for cardiovascular death by 10 years were 4.6% (SE = 2.0), 4.2% (SE = 1.2), and 3.3% (SE = 1.0), respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Despite substantial variability among individuals in the amount of depressive morbidity experienced during follow-up, the persistence of depressive symptoms was not associated with the risk of dying of cardiovascular causes. The consistency with which studies of community samples have linked depressive symptoms and cardiovascular mortality make this finding a surprising one because the processes presumed to underlie this linkage should be cumulative in their effects. Many inconsistencies exist in the relevant literature, though, and a number of methodological factors may be at play.

Some investigators who followed clinical samples have failed to find excess cardiovascular mortality (8–11). Most of these have determined the presence or absence of an excess in mortality through the use of age- and sex-specific vital statistics. In contrast, studies using community samples have compared subjects with more depressive symptoms to those with fewer symptoms by subsequent risks for cardiovascular death. The former approach requires more assumptions, some of which may not have been sufficiently valid in individual studies.

On the other hand, the success of efforts to statistically control for potentially confounding medical risk factors also rests on a number of assumptions. Perhaps the most tenuous of these involves the sensitivity with which risk factors are ascertained and the ways in which they are quantified. Aromaa et al. (4) simply considered whether heart disease or cerebral vascular disease were present at the beginning of follow-up. The studies of Barefoot and Schroll (2) and Everson et al. (3) considered a much wider array of risk factors and integrated them into continua for model testing. Unfortunately, no study design can completely eliminate the possibility of important confounds.

There may exist other, as yet unappreciated, physiological factors that predispose both to depressive illness and to cardiovascular morbidity. Such factors may raise risks for lifetime depressive disorder without bearing on the chronicity of depressive symptoms once the illness manifests. Consistent with this are data from the Epidemiological Catchment Area study, which revealed that individuals who had ever experienced 2 weeks of sustained dysphoria but who had never met criteria for MDD had a two-fold increase in risk of myocardial infarction in the next 6 years (5).

Threshold effects may underlie the failure to find a relationship between symptom persistence and cardiovascular mortality in this study. Because symptom severity and persistence are among the most powerful determinants of treatment seeking (29), subjects recruited through hospitals and clinics will, on average, exhibit longer episodes, more symptoms, and greater impairment than will depressed subjects identified in community samples. Patients ascertained at tertiary care centers, as those described here were, are likely to suffer from illness of even greater severity than are those treated by primary care providers. The link between depressive illness and cardiovascular mortality, however, may be more discernible at the milder end of the illness severity spectrum. Most studies of community samples have not attempted to identify severity thresholds below which the risk for subsequent cardiovascular death becomes negligible. It may well be that the least severe third described here overlaps substantially with groups identified as depressed in community samples (1, 2). Even patients described as "recovered" according to the CDS convention of 8 consecutive weeks with psychiatric status ratings of 1 or 2 have been shown to suffer from subsyndromal, but clinically meaningful, depressive symptoms during substantial portions of these "recovered periods" (30).

Several caveats and limitations in the current study warrant particular attention. First, the effects of depressive symptoms may be more malignant during periods of acutely increased cardiac vulnerability, and the results described here may, therefore, not be relevant to patients with recent cardiac events. Second, the ascertainment of baseline medical risk factors was relatively superficial and did not include physical examinations or laboratory testing. Third, the presence or absence of cardiovascular disease was ascertained through dichotomous responses to only one question, and the presence or absence of the important risk factors, obesity and smoking, was not determined at all. However, unrecognized associations between these risk factors and depressive morbidity during follow-up would have been more likely to produce a falsely positive association with cardiovascular death than a falsely negative one. Both smoking (31) and obesity (32–33) have been associated with the relative persistence of depressive symptoms and would be expected to artifactually produce associations between symptom chronicity and cardiovascular mortality.

The analyses presented here included no effort to estimate excess mortality relative to the general population. As noted in a recent and comprehensive review of the literature concerning mortality and depressive disorders, such an approach is quite liable to hidden but powerful biases (34). These arise from well-known associations between depressive disorders and certain cardiovascular risk factors, such as smoking, alcoholism, and obesity. The use of vital statistics to estimate expected mortality rates allows only for control of standard demographic variables.

The relevant literature does not allow firm conclusions regarding the role of depressive symptoms as long-term risk factors for cardiovascular mortality. Three studies of nonclinical samples suggest such a link, but one of these (3) concluded that hopelessness, rather than depressive symptoms generally, was the fundamental risk factor. Another adopted only superficial controls for potentially confounding risk factors (4). The results from clinical samples have been inconsistent, as noted previously. The present study used data from a high-surveillance, long-term follow-up to test the prediction that "more is worse," that is, that the persistence of depressive symptoms over time would correlate with eventual risks for cardiovascular death. No such relationship emerged. This suggests that previously described connections between depressive symptoms and long-term excesses in cardiovascular mortality reflect shared risk factors rather than a causal relationship between depressive illness and cardiovascular morbidity.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
All authors are from the National Institute of Mental Health Collaborative Program on the Psychobiology of Depression–Clinical Studies, which was conducted with the participation of the following investigators: M. B. Keller, MD (chairperson, Providence, RI); W. Coryell, MD (cochairperson, Iowa City, IA); J. D. Maser, PhD (Washington, DC); T. I. Mueller, MD, and M. T. Shea, PhD (Providence, RI); J. Fawcett, MD, and W. A. Scheftner, MD (Chicago, IL); W. Coryell, MD, and James Haley (Iowa City, IA); J. Endicott, PhD, A. C. Leon, PhD, and J. Loth, MSW (New York, NY); J. Rice, PhD, and T. Reich, MD (St. Louis, MO). Other contributors include H. S. Akiskal, MD, N. C. Andreasen, MD, PhD, P. J. Clayton, MD, J. Croughan, MD, R. M. A. Hirschfeld, MD, L. Judd, MD, M. M. Katz, PhD, P. W. Lavori, PhD, D. Solomon, MD, R. L. Spitzer, MD, and M. A. Young, PhD. Deceased contributors include G. L. Klerman, MD, E. Robins, MD, R. W. Shapiro, MD, and G. Winokur, MD. This manuscript was reviewed by the Publication Committee of the Collaborative Depression Study and has its endorsement.

Received for publication December 28, 1998.

Revision received July 7, 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 

1. Murphy JM, Monson RR, Olivier DC, Sobol AM, Leighton AH. Affective disorders and mortality. Arch Gen Psychiatry 1987; 44: 473–80.[Abstract/Free Full Text]
2. Barefoot JC, Schroll M. Symptoms of depression, myocardial infarction, and total mortality in a community sample. Circulation 1996; 93: 1976–80.[Abstract/Free Full Text]
3. Everson SA, Goldberg DE, Kaplan GA, Cohen RD, Pukkala E, Tuomilehto J, Salonen JT. Hopelessness and risk of mortality and incidence of myocardial infarction and cancer. Psychosom Med 1996; 58: 113–21.[Abstract/Free Full Text]
4. Aromaa A, Raitasalo R, Reunanen A, Impivaara O, Heiovaara M, Knect P, Lehtinen V, Joukamaa M, Maatela J. Depression and cardiovascular diseases. Acta Psychiatr Scand 1994; 377 (Suppl): 77–82.
5. Pratt LA, Ford DE, Crum RM, Armenian HK, Gallo JJ, Eaton WW. Depression, psychotropic medication, and risk of myocardial infarction. Circulation 1996; 94: 3123–9.[Abstract/Free Full Text]
6. Allgulender C. Suicide and mortality patterns in anxiety neurosis and depressive neurosis. Arch Gen Psychiatry 1994; 51: 708–12.[Abstract/Free Full Text]
7. Weeke A, Juel K, Vaeth M. Cardiovascular death and manic-depressive psychosis. J Affect Disord 1987; 13: 287–92.[Medline]
8. Ahrens B, Oerlinghausen B, Schou M, Wolf T, Aldo M, Grof E, Grof P, Lenz G, Simhandl C, Thau K, Vestergaard P, Wolf R, Moller JH. Excess cardiovascular and suicide mortality of affective disorders may be reduced by lithium prophylaxis. J Affect Disord 1995; 33: 67–75.[Medline]
9. Coryell W, Noyes R, Clancy J. Excess mortality in panic disorder. Arch Gen Psychiatry 1982; 39: 701–3.[Abstract/Free Full Text]
10. Tsuang MT, Woolson RF, Fleming JA. Causes of death in schizophrenia and manic depression. Br J Psychiatry 1980; 136: 239–42.[Abstract/Free Full Text]
11. Zilber N, Schufman N, Lerner Y. Mortality among psychiatric patients—the groups at risk. Acta Psychiatr Scand 1989; 79: 248–56.[Medline]
12. Frasure-Smith N, Lesperance F, Talajic M. Depression following myocardial infarction. JAMA 1993; 270: 1819–25.[Abstract/Free Full Text]
13. Frasure-Smith N, Lesperance F, Talajic M. Depression and 18-month prognosis after myocardial infarction. Circulation 1995; 91: 999–1005.[Abstract/Free Full Text]
14. Simonsick EM, Wallace RB, Blazer DG, Berkman LF. Depressive symptomatology and hypertension-associated morbidity and mortality in older adults. Psychosom Med 1995; 57: 427–36.[Abstract/Free Full Text]
15. Krittayaphong R, Cascio WE, Light KC, Sheffield D, Golden RN, Finkel JB, Glekas G, Koch GG, Sheps DS. Heart rate variability in patients with coronary artery disease: differences in patients with higher and lower depression scores. Psychosom Med 1997; 59: 231–5.[Abstract/Free Full Text]
16. Robinson RG, Starr LB, Price TR. A two-year longitudinal study of mood disorders following stroke. Br J Psychiatry 1984; 144: 256–62.[Abstract/Free Full Text]
17. Krishnan KR, Hays JC, Blazer DG. MRI-defined vascular depression. Am J Psychiatry 1997; 154: 497–501.[Abstract]
18. Vaillant GE, Orav J, Meyer SE, Vaillant LM, Roston D. Late-life consequences of affective spectrum disorder. Int Psychogeriatr 1996; 8: 13–32.
19. Glassman AH, Shapiro PA. Depression and the course of coronary artery disease. Am J Psychiatry 1998; 155: 4–11.[Abstract/Free Full Text]
20. Laghrissi-Thode F, Wagner WR, Pollock BG, Johnson PC, Finkel MS. Elevated platelet factor 4 and beta-thromboglobulin plasma levels in depressed patients with ischemic heart disease. Biol Psychiatry 1997; 42: 290–5.[Medline]
21. Musselman DL, Tomer A, Manatunga AK, Knight BT, Porter MR, Casey S, Marzec U, Harker LA, Nemeroff CB. Exaggerated platelet reactivity in major depression. Am J Psychiatry 1996; 153: 1313–7.[Abstract/Free Full Text]
22. Spitzer RL, Endicott J, Robins E. Research diagnostic criteria: rationale and reliability. Arch Gen Psychiatry 1978; 35: 773–82.[Abstract/Free Full Text]
23. Endicott J, Spitzer RL. A diagnostic interview: the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry 1978; 35: 837–44.[Abstract/Free Full Text]
24. Keller MB, Lavori PW, Friedman B, Neilsen E, Endicott J, McDonald-Scott P, Andreasen NC. The Longitudinal Interview Follow-Up Evaluation. Arch Gen Psychiatry 1987; 44: 540–8.[Abstract/Free Full Text]
25. Keller MB, Klerman GL, Lavori PW, Fawcett JA, Coryell W, Endicott J. Treatment received by depressed patients. JAMA 1982; 248: 1848–55.[Abstract/Free Full Text]
26. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–81.
27. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: Wiley and Sons; 1980.
28. Peto R, Peto J. Asymptotically efficient rank invariant procedures. J R Stat Soc 1972; 135: 185–206.
29. Coryell W, Endicott J, Winokur G, Akiskal H, Solomon D, Leon A, Mueller T, Shea T. Characteristics and significance of untreated major depressive disorder. Am J Psychiatry 1995; 152: 1124–9.[Abstract/Free Full Text]
30. Judd LL, Akiskal HS, Maser JD, Zeller PJ, Endicott J, Coryell W, Paulus MP, Kunovac JL, Leon AC, Mueller TI, Rice JA, Keller MB. A prospective 12-year study of subsyndromal and syndromal depressive symptomatology in unipolar major depressive disorders. Arch Gen Psychiatry 1998; 55: 694–701.[Abstract/Free Full Text]
31. Glassman AH, Helzer JE, Covey LS, Cuttler LB, Stetner F, Tipp JE, Johnson J. Smoking, smoking cessation, and major depression. JAMA 1990; 264: 1546–9.[Abstract/Free Full Text]
32. Black DW, Golstein RB, Mason EE. Prevalence of mental disorder in eighty-eight morbidly obese bariatric clinic patients. Am J Psychiatry 1992; 149: 227–34.[Abstract/Free Full Text]
33. Black DW, Golstein RB, Mason EE, Bell SE, Bloum N. Depression and other mental disorders in the relatives of morbidly obese patients. J Affect Disord 1992; 25: 91–6.[Medline]
34. Wulsin LR, Vaillant GE, Wells VE. A systematic review of the mortality of depression. Psychosom Med 1999; 61: 6–17.[Abstract/Free Full Text]

This article has been cited by other articles:

				J. G. Fiedorowicz, D. A. Solomon, J. Endicott, A. C. Leon, C. Li, J. P. Rice, and W. H. Coryell Manic/Hypomanic Symptom Burden and Cardiovascular Mortality in Bipolar Disorder Psychosom Med, July 1, 2009; 71(6): 598 - 606. [Abstract] [Full Text] [PDF]

				R. Morriss and F. A. Mohammed Metabolism, lifestyle and bipolar affective disorder J Psychopharmacol, November 1, 2005; 19(6_suppl): 94 - 101. [Abstract] [PDF]

				X. Zhang, S. L. Norris, E. W. Gregg, Y. J. Cheng, G. Beckles, and H. S. Kahn Depressive Symptoms and Mortality among Persons with and without Diabetes Am. J. Epidemiol., April 1, 2005; 161(7): 652 - 660. [Abstract] [Full Text] [PDF]

				L. R. Wulsin Does Depression Kill? Arch Intern Med, June 26, 2000; 160(12): 1731 - 1732. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Coryell, W. Articles by Keller, M. Search for Related Content PubMed PubMed Citation Articles by Coryell, W. Articles by Keller, M. Related Collections Depression Other Cardiovascular Medicine