Current clinical guidelines recommend that "elderly patients of both genders should be strongly encouraged to participate in exercise-based cardiac rehabilitation," 1^(p103) and small clinical trials have demonstrated the safety and efficacy of cardiac rehabilitation for improving exercise capacity and health status among older persons with heart disease. 2,3 Although older individuals, on the average, have poorer functional status than younger persons both before and after coronary artery bypass surgery, they experience a similar functional benefit from surgery. 4 Nonetheless, participation rates in cardiac rehabilitation programs, which can provide added benefits, are lower among older persons. 5

In this issue of the JCR, Dolansky and Moore 6 present the results of an observational study that evaluated the effect of participation in a cardiac rehabilitation program on lower extremity function and physical and mental health among persons 70 years of age or older after coronary artery bypass surgery. The results support another potential benefit of cardiac rehabilitation that is particularly relevant to this age group: improved lower extremity function. Exercise-based interventions have been shown to improve lower extremity function in several important subgroups of older persons, 7 so a similar benefit among individuals receiving cardiac rehabilitation follows logically. Because lower extremity function is one of the strongest predictors of future disability among older persons, 8 cardiac rehabilitation may not only improve exercise capacity, but also may reduce disability.

Using an observational design, Dolansky and Moore 6 provide evidence that participants in a cardiac rehabilitation program have better lower extremity function and perceived physical function than nonparticipants 6 months after hospital discharge. Although randomized controlled trials are considered the gold standard for evaluating the effectiveness of treatments, observational studies often can provide useful information. They also have certain advantages. For example, participants in observational studies often are more representative of the desired target population than participants in randomized controlled trials. Observational studies also permit the evaluation of treatment effectiveness in the context of "real-world" clinical practice rather than a potentially artificial research setting.

The main disadvantage of observational studies, as compared with randomized controlled trials, is their inability to account fully for the potential biases that typically arise in the absence of randomized assignment to treatment groups. Several methods have been proposed to overcome this potential bias including careful selection of the inclusion and exclusion criteria, statistical adjustment for potential confounders, and the use of masking in the ascertainment of outcomes. 9

The authors of the current study appropriately excluded participants who could not participate in cardiac rehabilitation, including those with significant neurologic or joint disease. However, the authors apparently did not ensure that the research staff who performed the outcomes assessments were masked to the participants' treatment group. Such masking is particularly important for outcomes such as lower extremity function, for which the assessments are effort dependent and potentially subjective.

Despite the use of several exclusion criteria, the two treatment groups exhibited differences that, although not statistically significant because of the small numbers, appear to be clinically meaningful. For example, only 6.2% of the cardiac rehabilitation participants had deficits in instrumental activities of daily living before surgery, and only 19% were New York Heart Association class 3 or 4, as compared to 18.2% and 36% of the nonparticipants, respectively. Because of the small sample size, the authors were somewhat limited in the number of potential confounders for which they could control. Nonetheless, the choice of potential confounders need not have been restricted to those that achieved statistical significance in the bivariate analyses. The authors' power calculations indicated a need for 100 participants to detect a moderate effect size with sufficient power, but in the end they were able to collect data for only 65 participants. This small sample size likely reflects the low participation rate in cardiac rehabilitation programs among older persons and the substantial obstacles to recruiting older individuals for research studies.

Despite the relatively small sample size, statistically significant differences favored the cardiac rehabilitation participants for all but 2 of the 7 outcomes after adjustment for a small number of potential confounders. Because the outcome measures are continuous and rely on possibly unfamiliar scales, readers may have some difficulty determining the clinical relevance of the differences found. Using Cohen's classification of effect sizes, 10 the study showed that cardiac rehabilitation had a small effect (.20 to .49) on muscle strength, ankle range of motion, and gait; a moderate effect (.50 to .79) on dynamic balance; and a large effect (>=.80) on perceived physical function. Descriptions of what mean differences might represent clinically offer an alternative strategy for gauging importance, one made possible by the authors' thorough descriptions of their measurement methods. For example, the mean difference between treatment groups in static balance was approximately 7, indicating the ability to stand on toes or on one foot for an additional 7 seconds. The mean difference between groups in dynamic balance was about 1, which corresponds to the difference between using one's arms and not using one's arms in standing from a chair (all other items remaining equal). 11 Although small, these differences may be clinically meaningful.

Whereas lower extremity function is a strong predictor of functional decline in older persons, the ultimate goal of cardiac rehabilitation is to improve outcomes that are important to patients. For older persons, a primary goal often is to restore and maintain functional independence rather than to increase longevity. The current study confirmed a statistically significant effect of cardiac rehabilitation on physical health status, as measured by the physical subscale of the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36). However, the clinical interpretation of differences in SF-36 scores can be difficult. In previous studies, cardiac rehabilitation has been shown to result in an 8-point improvement in the physical component of the SF-36, which is similar to the 7.5 point mean difference between groups detected in the current study. 12,13 Larger studies would be needed to assess the effect of cardiac rehabilitation on clinically relevant functional outcomes such as disability in mobility and activities of daily living.

Outcomes in this study were assessed 6 months after hospital discharge, approximately 6 weeks after the end of the cardiac rehabilitation programs. Previous studies of cardiac rehabilitation for older persons have shown that benefits seen shortly after completion of the program often are lost 12 months after surgery. 2,3 One randomized controlled trial comparing center-based cardiac rehabilitation, home-based rehabilitation, and no rehabilitation among postmyocardial infarction patients demonstrated that although both rehabilitation programs improved total exercise capacity and quality of life among individuals older than 75 years, only the home-based group had maintained their gains at 1 year. 3 To evaluate the potential benefits of cardiac rehabilitation more completely, future research should consider follow-up periods longer than 6 months.

The current study provides preliminary evidence that participation in a cardiac rehabilitation program improves lower extremity function in older persons after coronary artery bypass surgery. Additional studies are needed to confirm this finding and to determine whether this improvement in lower extremity function ultimately leads to a reduction in disability. Although this question would be best addressed in a randomized controlled trial, it also could be evaluated in a well-designed observational study with blinded assessments of outcomes, an extended follow-up period, and a sample size large enough to account adequately for likely differences in potential confounders between treatment groups and to provide sufficient power for detecting clinically meaningful between-group differences. According to the current evidence of effectiveness and safety, greater efforts should be made to encourage older individuals to participate in cardiac rehabilitation programs after coronary artery bypass surgery.

References