Article Text

Download PDFPDF

Psychotropic medication use and risk of adverse cardiovascular events in women with suspected coronary artery disease: outcomes from the Women’s Ischemia Syndrome Evaluation (WISE) study
  1. D S Krantz1,
  2. K S Whittaker1,
  3. J L Francis1,
  4. T Rutledge2,
  5. B D Johnson3,
  6. G Barrow3,
  7. C McClure3,
  8. D S Sheps4,
  9. K York4,
  10. C Cornell5,
  11. V Bittner6,
  12. V Vaccarino7,
  13. W Eteiba3,
  14. S Parashar7,
  15. D A Vido8,
  16. C N Bairey Merz9
  1. 1
    Department of Medical and Clinical Psychology, Uniformed Services University of the Health Sciences, Bethesda, Maryland, USA
  2. 2
    VA San Diego Healthcare System, San Diego, California, USA
  3. 3
    Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
  4. 4
    Department of Medicine, Division of Cardiology, University of Florida, Gainesville, Florida
  5. 5
    Department of Health Behavior and Health Education, University of Arkansas for Medical Sciences, Little Rock, Arkansas, USA
  6. 6
    University of Alabama at Birmingham, Alabama, USA
  7. 7
    Emory University, Atlanta, Georgia, USA
  8. 8
    Allegheny General Hospital, Pittsburgh, Pennsylvania, USA
  9. 9
    Heart Institute, Cedars-Sinai Research Institute, Cedars-Sinai Medical Center, Los Angeles, California, USA
  1. Correspondence to Dr David S Krantz, Uniformed Services University of the Health Sciences, Department of Medical and Clinical Psychology, 4301 Jones Bridge Road, Bethesda, MD 20814, USA; dskrantz{at}


Objective: This study investigated the relation between psychotropic medication use and adverse cardiovascular (CV) events in women with symptoms of myocardial ischaemia undergoing coronary angiography.

Method: Women enrolled in the Women’s Ischemia Syndrome Evaluation (WISE) were classified into one of four groups according to their reported antidepressant and anxiolytic medication usage at study intake: (1) no medication (n = 352); (2) anxiolytics only (n = 67); (3) antidepressants only (n = 58); and (4) combined antidepressant and anxiolytics (n = 39). Participants were followed prospectively for the development of adverse CV events (for example, hospitalisations for non-fatal myocardial infarction, stroke, congestive heart failure and unstable angina) or all-cause mortality over a median of 5.9 years.

Results: Use of antidepressant medication was associated with subsequent CV events (HR 2.16, 95% CI 1.21 to 3.93) and death (HR 2.15, 95% CI 1.16 to 3.98) but baseline anxiolytic use alone did not predict subsequent CV events and death. In a final regression model that included demographics, depression and anxiety symptoms, and risk factors for cardiovascular disease, women in the combined medication group (that is, antidepressants and anxiolytics) had higher risk for CV events (HR 3.98, CI 1.74 to 9.10, p = 0.001 and all-cause mortality (HR 4.70, CI 1.7 to 2.97, p = 0.003) compared to those using neither medication. Kaplan-Meier survival curves indicated that there was a significant difference in mortality among the four medication groups (p = 0.001).

Conclusions: These data suggest that factors related to psychotropic medication such as depression refractory to treatment, or medication use itself, are associated with adverse CV events in women with suspected myocardial ischaemia.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Depression is an independent risk factor for cardiac events in patients with coronary artery disease (CAD)1 2 and may increase cardiovascular (CV) risk in healthy populations.3 Anxiety may also be a CAD risk factor.2 Therefore, studies have examined the efficacy of psychotropic medications, particularly selective serotonin reuptake inhibitors (SSRIs), in CAD patients with depressive symptoms.

Several randomised trials have demonstrated short-term safety of SSRIs in CAD patients. The CREATE study reported no 12-week increased risk of cardiovascular events for citalopram in CAD patients.4 The SADHART study of patients with acute myocardial infarction (MI) or unstable angina, reported that sertraline, compared to placebo, did not affect left ventricular (LV) function, and other important CV parameters5 over a 24-week period.

In the ENRICHD trial6 of cognitive behavioural therapy (CBT) to reduce morbidity/mortality in CAD patients with depression and/or low social support, many patients in the study were treated with, but not randomly assigned to, SSRIs. Secondary analysis of these data showed that SSRI use was associated with reduced non-fatal MI, and cardiac and all-cause mortality.7 Two other observational studies have also found fewer adverse CV effects in patients taking SSRIs.8 9

Use of tricyclic antidepressants (TCA) has been associated with increased risk of malignant ventricular arrhythmias and sudden cardiac death.10 11 One retrospective study12 found that TCA use was associated with a dose-related increase in sudden cardiac death. Despite potential negative effects associated with many TCAs, low doses of imipramine have been shown to be effective in treating patients with unexplained chest pain, and low TCA doses do not appear to be associated with increased risk.13

Although causal relations between antidepressant use and adverse CV events can only be made from randomised clinical trials, four other cohort studies are provocative3 8 14 15 in showing a positive association between antidepressant use and increased CV events and/or mortality in diverse populations of CAD patients,3 15 16 even after rigorous control for pre-existing depression. Most recently, in a cohort of over 63 000 women initially free of CAD, Whang et al observed that antidepressant medication use with depression was associated with a more than threefold risk of sudden cardiac death and fatal CHD over a 3½-year follow-up period.3

The effects of anxiolytic medications in CAD patients have primarily been investigated in observational studies. In a 10-year follow-up of 500 men,17 anxiolytic and sedative medication use was associated with increased CV and all-cause mortality. Another observational study18 found that benzodiazepine use was associated with increased CV event risk. These studies did not statistically control for levels of anxiety or depression that could have been responsible for the observed increase in CV risk. Anxiety and depressive disorders commonly co-occur, and medication studies would be improved by jointly addressing both anxiolytic and antidepressant psychotropic treatments.

The issue of psychotropic medication use may be particularly relevant to women with CAD since psychotropic medication is more prevalent in women than men.19 The issue of psychotropic medication use may be particularly relevant to women with CAD since psychotropic medication is more prevalent in women than men. The Women’s Ischemia Syndrome Evaluation (WISE) is a National Heart, Lung and Blood Institute-sponsored study designed to address the diagnosis of ischaemic heart disease and mechanisms of ischaemia in women with suspected CAD. The purpose of the present report is to describe the relation between psychotropic medication use and adverse CV events in women undergoing coronary angiography for suspected myocardial ischaemia. To rigorously account for confounding effects of previous symptoms, the present analyses control for levels of depression and anxiety and also control for previous depression treatment history. Because of the co-morbidity of anxiety and depressive disorders, the combined use of antidepressants and anxiolytic medications is often observed. Therefore, the present study also investigates associations of incident CV events with the combined medications, as well as individual use of these medications.


Participants were enrolled in the WISE study, a collaborative study designed to optimise diagnostic evaluation and testing for ischaemic heart disease in a cohort of 936 women, using a study design and methods previously published.20 To be enrolled in WISE, women had to be at least 18 years of age and be undergoing clinically indicated coronary angiography. Exclusion criteria included co-morbidity compromising one-year follow-up, pregnancy, contraindications to provocative diagnostic testing, cardiomyopathy, NYHA class IV congestive heart failure, recent myocardial infarction, significant valvular or congenital heart disease and inability to complete questionnaires. The present study included 519 WISE participants both with and without angiographic CAD who had complete psychosocial data and psychotropic medication information.

The initial WISE visit consisted of a physical examination, medical history and demographic, symptom, psychosocial assessment; all participants also underwent a coronary angiogram, subsequently analysed by investigators blind to all WISE clinical data. Obstructive CAD was defined as the presence of ⩾50% stenosis in ⩾1 epicardial coronary artery, and a continuous CAD severity score was assessed, ranging from 5 to 88.5. Psychotropic medication usage was obtained from medication information collected during a baseline interview. Women were asked about treatment received within the last 6 weeks before study entry for (1) “antidepressants” and (2) “anxiolytics, sedatives or hypnotics”. Response choices were “yes”, “no” or “unknown”. Specific subclasses of antidepressants or anxiolytics were not assessed. The Beck Depression Inventory (BDI), a well-validated 21-item questionnaire was used to assess depressive symptomatology. Trait anxiety was assessed using the 10-item trait anxiety scale from the validated State Trait Personality Inventory. Higher scores on the BDI and the State Trait Anxiety Inventory (STAI) scale indicate greater levels of depression and anxiety symptoms, respectively.

Follow-up for the occurrence of CV events was obtained by telephone interview at 6 weeks and yearly thereafter for occurrence of major cardiovascular events, including death due to CV or hospital stays for non-fatal MI, stroke or congestive heart failure. Deaths were verified via death certificate with cause blindly reviewed by an events committee. Median length of follow-up was 5.9 years. The primary outcome variables for this study were incidence of adverse CV events and all-cause mortality, which were examined separately. For the purposes of this study, women were classified into four groups according to their antidepressant and anxiolytic medication status during the week before intake: (1) no medication (n = 353); (2) anxiolytics only (n = 68); (3) antidepressants only (n = 59); and (4) combined antidepressants and anxiolytics (combined medication; n = 39).

Variables are expressed as means (SD) or as percentages. Group differences were assessed by non-parametric Kruskal-Wallis χ2 tests. Age-adjusted Cox regression analyses and stepwise Cox regression analyses were used to assess time to adverse CV events and all-cause mortality among the four medication groups. The no medication group was used as the reference category. Age-adjusted Kaplan-Meier survival curves were plotted to examine the association of medication group with mortality and adverse CV events. The logrank test was used to compare survival between the medication groups. The criterion for significance was set at p<0.05. Analyses were conducted using SAS version 8.2.


Demographic and clinical variables

Demographic and clinical characteristics for the four medication groups are presented in table 1.

Table 1

Demographic, clinical characteristics and cardiovascular (CV) outcomes by medication group (n = 519)

Compared to women not taking medication, women in the combined medication groups were more likely to have ever smoked and to have a history of smoking, hypertension, dyslipidaemia and non-fatal MI. Women in the combined group also had greater depression and anxiety scores.

Risk factors and psychotropic medication

During the median follow-up of 5.9 years, there were a total of 81 (15.6%) new CV events and 46 (8.8%) all-cause deaths. CV events (and percentage) and total mortality for each of the four medication groups are presented in table 1. In unadjusted analyses, baseline use of antidepressant medication was associated with subsequent cardiovascular events (hazard ratio (HR) 2.16, 95% CI 1.21 to 3.93) and death (HR 2.15, 95% CI 1.16 to 3.98). The associations between baseline anxiolytic use and subsequent cardiovascular events and death were not significant.

Risk factors, medical history variables and the medication use categories were entered into multivariate analyses that also included initial depression and anxiety scores. In the full multivariate Cox regression model (table 2), combined medication use was a significant predictor of both CVD events and mortality, even after adjusting for baseline BDI and STAI scores (p = 0.001 and 0.003, respectively). Additional significant covariates for both CV events and total mortality were higher education and CAD severity score, as well as a history of smoking and diabetes history. Total mortality was related to age, CAD severity, smoking and diabetes history. In addition BDI scores were positively correlated with adverse CV events and total mortality (p = 0.02 and p = 0.01, respectively), and STAI scores were inversely related to CV events and total mortality (p = 0.02 and 0.002, respectively).

Table 2

Cox regression model for risk factors and cardiovascular (CV) events and mortality

Age-adjusted Kaplan-Meier survival analyses for all-cause mortality (see figs 1, 2) indicated a significant difference in mortality and CV events among the medication groups. In both models, the combined medication group experienced higher CV event rates compared to the no medication group (logrank (Mantel Cox) test for all-cause mortality = 6.9 (χ2, 1 df), p = 0.009; log rank test for CV events = 16.5 (χ2, 1 df), p<0.001.

Figure 1

Event-free survival curves for all cause mortality by medication group.

Figure 2

Event-free survival curves for cardiovascular events by medication group.

In age-adjusted regression models, combined medication use was associated with increased risk for CV events (HR 2.80, 95% CI 1.70 to 4.80) and all-cause mortality (HR 2.60, 95% CI 1.40 to 5.00). The final multivariate Cox regression models presented in table 2 indicate that the combined medication group had a higher risk for CV events (p = 0.001) and all-cause mortality (p = 0.003).

History of depression treatment

We also examined history of depression treatment as an additional covariate (with BDI scores remaining in the model), since it overlaps considerably with psychotropic medication use and has been associated with increased cardiovascular risk in WISE.21 Despite the shared variance between BDI scores and depression treatment history, when history of depression treatment was added as a covariate in the regression analyses, the relation between combined medication use and adverse CV events was attenuated but remained statistically significant (HR 2.7, 95% CI 1.10 to 6.90, p = 0.04). However, all-cause mortality (HR 2.86, 95% CI 0.82 to 9.96, p = 0.10) was no longer significant.


This is one of several prospective cohort studies to observe an association between psychotropic medication use and adverse CV outcomes in both patients with CAD and in an initially healthy population.3 15 16 22 The study is unique, however, in demonstrating this relation in a population of symptomatic women with suspected CAD, and these findings may be particularly relevant since psychotropic medication use is more prevalent in women than men,23 and these medications are frequently prescribed for women with chest pain and no angiographic evidence of CAD. Baseline use of antidepressant medication was associated with subsequent cardiovascular events and death in univariate analyses. Although we did not observe such a relation with anxiolytics alone, in this cohort of women with suspected myocardial ischaemia and CAD, in multivariate analyses, the often observed joint use of antidepressants and anxiolytics was associated with increased CV event and total mortality risk.

Previous studies of CAD patients have examined heart failure patients,15 CAD patients undergoing coronary artery bypass grafting16 and predominantly male patients undergoing cardiac catheterisation.24 The present study extends these observations to the important population of women with symptoms of ischaemia and suspected CAD. Another recent study of more than 63 000 women enrolled in the Nurses Health Study also observed that reported antidepressant use (which included both SSRIs and other antidepressants) in a population unselected for presence of CAD was prospectively associated with incidence sudden cardiac death, fatal MI and non-fatal CHD events.3

Although the present results are consistent with three additional long-term observational cohort studies,3 8 9 these studies are in contrast to two short-term clinical trials of SSRIs in depressed cardiac patients5 7 that did not demonstrate an association of these antidepressant medications with increased CV events. A post hoc analysis of clinical trial data found a beneficial relation between SSRI use and cardiac outcomes.4 The limitations and perils of drawing conclusions from observational studies as opposed to the definitive standard of clinical trials are well known, and have been discussed in relation to studies of antidepressants and cardiac outcomes.14 In addition, the present study only obtained data on medication use at study entry and it therefore cannot be determined whether patients are taking psychotropics at time of death or CVD event. Based on these observations, a definitive assessment of the effects of psychotropic medications on morbidity/mortality in women with suspected CAD and other CAD populations requires additional investigation.

Psychotropic medication use is probably a proxy for severity of psychiatric symptoms, treatment-refractory and/or residual depression in this sample, and/or other comorbidity, and has been used as such in previous research.3 Approximately 30% of the women were taking an anxiolytic, antidepressant, or both, in this study yet these patients remained with higher levels of symptomatology than women not taking medication. It has been previously demonstrated25 that depression, and, to a lesser extent, anxiety26 are independent risk factors for CV disease. In non-patient populations, dual diagnoses of generalised anxiety disorder and major depressive disorder may be associated with more significant impairment.27 The possibility that depressive and anxiety symptoms may be linked to several of the same mechanisms involved in the development of CAD is currently under investigation, with some studies finding no evidence for a joint effect,2 while other results suggest that there may be a combined effect of anxiety and depression on cardiovascular outcomes.28 Thus, there is a possibility that symptoms of depression and anxiety may be driving the elevated risk for adverse CV outcomes in these women, with the highest risk being evident in women taking both antidepressants and anxiolytic medications, and therefore those with more severe psychological disturbance. Supporting this possibility is a previous report from the WISE study21 demonstrating that depressive symptoms and reported treatment history for depression are predictive of mortality. However, in our analyses we statistically accounted for symptomatology by including the BDI and anxiety measures as covariates. Risk for all-cause mortality and CV events was attenuated but remained significant for the combined medication group even after controlling for psychiatric symptoms. Adding history of depression treatment as an additional covariate similarly did not eliminate the effect of combined medications on adverse CV events, despite the fact that treatment history is itself a strong proxy variable for psychotropic medication use.

Study limitations

A limitation of this study, as in other observational studies of antidepressant use and CAD outcomes, is that women with psychiatric symptoms were not randomly assigned to medication treatment. Further, medication usage was self-reported. Thus, a causal relation between psychotropic medications and CV events cannot be established. Only a randomised clinical trial of psychotropic medications adequately powered to detect adverse CV events can assess causal relations and rule out confounding factors, such as severity of psychiatric symptoms. Another study limitation is the fact that the present medication assessments did not distinguish between SSRI and tricyclic antidepressants; although the latter are linked with possible adverse CV events, they may have been used for chest pain of unknown origin.13 Regardless of prescription pattern of particular antidepressant and anxiolytic medications in these women with suspected CAD, it is of note that those patients who were prescribed these medications were, for a variety of possible reasons, at increased risk of subsequent CV events. The present results were obtained after rigorous statistical controls for two indices of previous psychopathology—both depression and anxiety symptoms and also psychiatric treatment history. Despite these rigorous statistical controls, it is possible that the BDI and trait anxiety measures do not adequately capture the type of psychological distress typically experienced by patients with undiagnosed chest pain. For example, high rates of panic disorder29 are found in patients referred for cardiac examination, and panic and/or phobic anxiety may be related to increased cardiac death.26 Thus, controlling for symptoms such as trait anxiety and depressive symptoms may not be adequate, and treatment with psychotropic medications may not eliminate the increased risk in this sample associated with psychiatric symptoms.

Clinical implications

The diagnosis of CAD in women presents particular challenges, and many women experience depression, anxiety and impaired quality of life because of persistent symptoms.30 Psychotropic medications may be prescribed for these indications. The present data suggest that factors related to psychotropic medications, or medication use itself, are associated with increased CV events in women with suspected myocardial ischaemia. It is unknown whether the depression and anxiety symptoms reported in this sample reflect under-treatment of psychiatric symptoms, a poor response to treatment, a treatment bias effect or some other factor. Regardless, the persistence of psychiatric distress in women with chest pain and suspected ischaemia needs to be addressed clinically. These symptoms are not only associated with elevated CV event risk but also with impaired psychological functioning31 and significant economic burden32 and therefore should be addressed clinically. Future research needs to focus on identifying associations between use of psychotropic medications for depression and anxiety and adverse CV events in women with myocardial ischaemia in prospective, controlled clinical trials.



  • Funding This work was supported by contracts from the National Heart, Lung and Blood Institutes, nos N01-HV-68161, N01-HV-68162, N01-HV-68163, N01-HV-68164, grants U0164829, U01 HL649141, U01 HL649241, T32HL69751, and grants from the Gustavus and Louis Pfeiffer Research Foundation, Danville, NJ, The Women’s Guild of Cedars-Sinai Medical Center, Los Angeles, CA, The Ladies Hospital Aid Society of Western Pennsylvania, Pittsburgh, PA, and QMED, Inc, Laurence Harbor, NJ, the Edythe L Broad Women’s Heart Research Fellowship, Cedars-Sinai Medical Center, Los Angeles, California, and the Barbra Streisand Women’s Cardiovascular Research and Education Program, Cedars-Sinai Medical Center, Los Angeles.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • See Editorial, p 1893

Linked Articles