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


Meta-analysis: Psychosocial interventions augment cardiac rehabilitation programs

ACP J Club. 1996 Sept-Oct;125:34. doi:10.7326/ACPJC-1996-125-2-034

Source Citation

Linden W, Stossel C, Maurice J. Psychosocial interventions for patients with coronary artery disease. A meta-analysis. Arch Intern Med. 1996 Apr 8; 156:745-52.



To determine whether the addition of psychosocial interventions to standard cardiac rehabilitation programs improves mortality, nonfatal cardiac event recurrences, psychological distress (anxiety and depression), and clinical risk factors in coronary heart disease (CHD).

Data sources

MEDLINE searches and bibliographies of review articles and studies were used to identify randomized controlled trials.

Study selection

Trials were selected if they included patients who were in cardiac rehabilitation programs and had confirmed CHD at randomization. Patients in the control arms of the trials had to receive usual care and patients in the experimental arms had to receive usual care plus psychosocial treatment. Treatments considered were group psychotherapy, relaxation, individual therapy, cognitive-behavioral therapy, stress management, individual psychological support as needed, music therapy, and group education.

Data extraction

Data were extracted on the type of intervention and therapists involved, mortality, recurrences of CHD, anxiety, depression, blood pressure, heart rate, cholesterol levels, and follow-up duration. Weighted log-adjusted odds ratios (ORs) were calculated for mortality and morbidity data. Pooled mean effect sizes (ESs) were calculated for control and experimental arms across studies, and differences between the effect sizes were reported; negative differences favored improved outcome with the psychosocial interventions.

Main results

23 studies were included; 12 studies evaluated mortality, and 10 evaluated recurrences. 2024 patients received psychosocial interventions, and 1156 patients received standard care. Patients who received psychosocial interventions showed reduced mortality at ≤ 2 years follow-up (10 studies) (OR 1.70,95% CI 1.09 to 2.64) and fewer recurrences of CHD at ≤ 2 years follow-up (8 studies) (OR 1.84, CI 1.12 to 2.99) and at > 2 years follow-up (3 studies) (OR 1.64, CI 1.06 to 2.54). Benefits of psychosocial interventions were also shown (negative ESs) for psychosocial distress (ES -0.34 {CI -0.55 to -0.13}*), systolic blood pressure (ES -0.24 {CI -0.35 to -0.13}*), heart rate (ES -0.38 {CI -0.5 to -0.26}*), and cholesterol levels (ES -1.54 {CI -1.91 to -1.25}*). No difference existed in diastolic blood pressure or mortality between the groups after 2 years.


The addition of psychosocial interventions to standard cardiac rehabilitation programs after CHD improves mortality, reduces recurrence of CHD, psychological distress (anxiety and depression), and some biological risk factors, especially within the first 2 years.

Sources of funding: British Columbia and Yukon Heart and Stroke Foundation and the British Columbia Health Research Foundation.

For article reprint: Dr. W. Linden, Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, BC, V6T 1Z4 Canada. FAX 604-822-6923.

*Numbers calculated from data in article.


Patients who have had myocardial infarction (MI) are at high risk for subsequent mortality. Cholesterol reduction, β-blockers, aspirin, and angiotensin-converting enzyme inhibitors have been shown to be beneficial for patients with CHD. Observational studies indicate that the prognosis after MI is adversely affected by depression (1) and lack of social support (2). Observational studies do not address whether these factors can be modified. The meta-analysis by Linden and colleagues may help to do so. Pooling studies that included various psychosocial interventions implemented by practitioners for groups, couples, or individuals show an important benefit. Most patients had had MI and, as such, were at high risk for subsequent CHD death or events. A statistically significant risk reduction of 40% to 45% occurred at least in the first 2 years. Applying this result to a typical group of patients with CHD, such as the placebo group in the 4S trial (3), approximately 30 persons would need to receive the psychosocial intervention to prevent 1 CHD death within 2 years. This effect is similar to that of simvastatin but is larger than that of β-blockers, aspirin, or cardiac rehabilitation alone; however, the 95% CIs include a number needed to treat of 140.

The other outcomes of this meta-analysis, standardized as "effect sizes," are more difficult to interpret. An ES of 1.0 is a difference of 1 SD. The mean ESs are modest and the changes in blood pressure and cholesterol may not explain the benefit. Improved compliance with diet, medication, and exercise may have played a role. The decrease in psychosocial distress may reflect an additional benefit by improving well-being. The authors raise the concern of the cost-effectiveness of these psychosocial interventions and suggest that patients with psychosocial difficulties during their rehabilitation should receive these interventions. A psychosocial history or a simple screening test might identify the patients who are most likely to benefit.

Craig Redfern, DO
Providence Portland Medical CenterPortland, Oregon, USA


1. Frasure-Smith N, Lespérance F, Talajic M. JAMA. 1993;270:1819-25.

2. Berkman LF. Psychosom Med. 1995; 57:245-54.

3. The Scandinavian Simvastatin Survival Study (4S). Lancet. 1994;334:1383-9.