Current issues of ACP Journal Club are published in Annals of Internal Medicine


Depressive symptoms predicted acute myocardial infarction and death

ACP J Club. 1996 Sept-Oct;125:52. doi:10.7326/ACPJC-1996-125-2-052

Source Citation

Barefoot JC, Schroll M. Symptoms of depression, acute myocardial infarction, and total mortality in a community sample. Circulation. 1996 May 15;93:1976-80.



To determine the association between depressive symptoms and risk for acute myocardial infarction and death.


27-year cohort analytic study.


Community in Glostrup, Denmark.


Eligible participants were residents of Glostrup who were born in 1914 and had taken part in an epidemiologic study of cardiovascular disease that involved physical examinations and psychological testing. Examinations were done in 1964 and 1974. 730 participants (56% men) had psychological testing measures available from at least 1 examination. Follow-up was > 99%.

Assessment of risk factors

The 40-item obvious depression (OBD) subscale of the Minnesota Multiphasic Personality Inventory was used to measure depressive symptoms. Risk factors considered were blood pressure, blood chemistry, smoking, physical activity, body mass, myocardial ischemia, and pulmonary function. 4 models were used to assess the ability of OBD scores to predict acute myocardial infarction and all-cause mortality. The first model controlled for age and sex; the second controlled for age, sex, and other risk factors for acute myocardial infarction or mortality; the third controlled for age and sex and was based on a reduced sample from which participants with possible myocardial ischemia at baseline had been eliminated; and the fourth was based on a reduced sample and excluded 9 items on the OBD scale about somatic complaints.

Main outcome measures

Association of OBD scores with acute myocardial infarction and all-cause mortality.

Main results

At the end of 1991, 122 participants (17%) had had an acute myocardial infarction and 290 participants (40%) had died. Ischemic heart disease and myocardial infarction accounted for 88 (30%) of deaths. Higher OBD scores were associated with increased risk for acute myocardial infarction in all 4 models. For a 2-standard deviation difference from the mean depression score, the relative risk (RR) for acute myocardial infarction was 1.70 (95% CI 1.23 to 2.34, P = 0.002) after controlling for traditional risk factors and was 1.50 (CI 1.02 to 2.19, P = 0.045) for the reduced sample and with somatic items omitted. For total mortality, the corresponding RRs were 1.59 (CI 1.26 to 2.00) and 1.35 (CI 1.06 to 1.75), respectively.


High scores on measures of depressive symptoms were associated with increased risks for acute myocardial infarction and death from all causes.

Sources of funding: Fogarty International Center and National Institutes of Health.

For article reprint: Dr. J.C. Barefoot, Box 2969, Duke University Medical Center, Durham, NC 27710, USA. FAX 919-681-8960.


It has long been recognized that psychological symptoms are associated with an increased risk for physical disorders (1). The study by Barefoot and Schroll is consistent with others that show that depression is associated with a subsequent increase in mortality from myocardial infarction, not only for psychiatric patients with major affective disorder (2) but for patients with a wide range of severity of depressive symptoms in community samples. The association between depression and mortality, both from myocardial infarction and from all other causes, persisted after evidence of pre-existing heart disease was controlled for and after somatic symptoms of depression, which can represent subclinical physical disease, were omitted from the depression ratings.

The association was independent of other risk factors. Smoking was positively correlated with depression among women. Another possible confounder that was not reported in the study is the use of antidepressant drugs that can adversely affect cardiac function. The increased coronary risk, however, was not confined to a group of high-risk patients with a specific psychiatric condition; rather, a graded risk was found across all 4 quartiles of depression scores. The authors conceptualize the risk factor in terms of a chronic psychological predisposition to myocardial infarction. The mechanism is unknown.

If persons with a depressive personality type are more prone to myocardial infarction, it does not necessarily follow that interventions that reduce depressive symptoms will decrease the risk for infarction. These findings suggest, however, that co-existing depression should prompt greater efforts to tackle other coronary risk factors and should heighten the physician's suspicions when he or she is deciding whether to investigate symptoms that may represent ischemic heart disease. No evidence indicates that regular testing of depressed patients for ischemic heart disease is cost-effective.

Tony Kendrick, MD
St. George's Hospital Medical SchoolLondon, England, UK


1. Eastwood MR, Trevelyan MH. Relationship between physical and psychiatric disorders. Psychol Med. 1972;2:363-72.

2. Tsuang MT, Woolson RF, Fleming JA. Premature death in schizophrenia and affective disorders. An analysis of survival curves and variables affecting the shortened survival. Arch Gen Psychiatry. 1980;37:979-83.