Determinants of physical activity after hospitalization for coronary artery disease: the Tracking Exercise After Cardiac Hospitalization (TEACH) Study

Abstract

Background

Little is known about physical activity levels in patients with coronary artery disease (CAD) who are not engaged in cardiac rehabilitation. We explored the trajectory of physical activity after hospitalization for CAD, and examined the effects of demographic, medical, and activity-related factors on the trajectory.

Design

A prospective cohort study.

Methods

A total of 782 patients were recruited during CAD-related hospitalization. Leisure-time activity energy expenditure (AEE) was measured 2, 6 and 12 months later. Sex, age, education, reason for hospitalization, congestive heart failure (CHF), diabetes, and physical activity before hospitalization were assessed at recruitment. Participation in cardiac rehabilitation was measured at follow-up.

Results

AEE was 1948 ± 1450, 1676 ± 1290, and 1637 ± 1486 kcal/week at 2, 6 and 12 months, respectively. There was a negative effect of time from 2 months post-hospitalization on physical activity (P<0.001). Interactions were found between age and time (P = 0.012) and education and time (P = 0.001). Main effects were noted for sex (men more active than women; P<0.001), CHF (those without CHF more active; P<0.01), diabetes (those without diabetes more active; P<0.05), and previous level of physical activity (those active before hospitalization more active after; P<0.001). Coronary artery bypass graft patients were more active than percutaneous coronary intervention (PCI) patients (P = 0.033).

Conclusions

Physical activity levels declined from 2 months after hospitalization. Specific subgroups (e.g. less educated, younger) were at greater risk of decline and other subgroups (e.g. women, and PCI, CHF, and diabetic patients) demonstrated lower physical activity. These groups need tailored interventions.

Keywords

coronary disease exercise follow-up studies physical activity prevention women

Introduction

Regular physical activity is safe for most individuals with coronary artery disease (CAD), and is associated with decreased mortality and improved quality of life [1].

Current recommendations suggest that CAD patients should accumulate at least 30 min of moderate-intensity physical activity on most days of the week [1], an energy expenditure of almost 1000 kcal/week [2]. Additional benefits may be associated with higher levels of energy expenditure; Hambrecht et al. [3] have suggested that 1500 kcal/week is necessary to improve cardiorespiratory fitness and stop disease progression, whereas 2200 kcal/week is required for disease regression.

Surprisingly little is known about typical physical activity levels in patients with CAD. What is known comes mostly from studies of individuals enrolled in structured cardiac rehabilitation programs [4–7]. Most patients enrolled in these program (≍70%) achieve the 1000 kcal/week level of weekly leisure-time activity energy expenditure (AEE), whereas 40–45% achieve the 1500 kcal/week level. However, physical activity data obtained from patients in cardiac rehabilitation may not be representative because 70–85% of patients with CAD do not participate in these programs [8, 9]. Existing rehabilitation-based studies of physical activity have generally included small numbers of women, older patients, and patients undergoing percutaneous coronary interventions (PCI) [10]. There is a need for additional studies of physical activity behavior in CAD patients that include both those that do and do not participate in cardiac rehabilitation.

The goals of the present study were: (i) to explore the trajectory of physical activity behavior in the year after hospitalization for CAD; and (ii) to evaluate the effect of sex, age group (≤ 65 or > 65 years), level of education, reason for hospitalization [acute myocardial infarction (AMI), PCI, or coronary artery bypass graft (CABG) surgery], congestive heart failure (CHF), diabetes, physical activity level before hospitalization, and participation in cardiac rehabilitation on the trajectory.

On the basis of previous studies showing decaying levels of physical activity over time during follow-up in cardiac rehabilitation patients [4, 7, 11], we hypothesized that there would be a negative trajectory for physical activity after hospitalization for CAD in general. We expected the trajectory to be moderated by variables such as participation in structured cardiac rehabilitation [7] and previous participation in physical activity [12]. We also anticipated lower physical activity levels in women, older patients, the less educated, and in patients with CHF or diabetes.

Methods

The Tracking Exercise After Cardiac Hospitalization (TEACH) Study is a prospective cohort study designed to examine patterns and predictors of physical activity behavior in patients with CAD. Participants were recruited while hospitalized for CAD and were followed for a period of 1 year. Participants completed a wide battery of assessments, including demographic, medical, behavioral, environmental and psychosocial factors. In this paper, we examine demographic, medical and behavioral data.

Subjects

Participants were between the ages of 20 and 85 years and had been hospitalized for AMI, PCI, or CABG at cardiac centers in Ottawa (two sites) and Kingston (one site), Canada. Patients with contraindications to exercise (i.e. unstable angina; uncontrolled cardiac arrhythmias; neuromuscular, musculoskeletal, or rheumatoid disorders exacerbated by exercise; uncontrolled diabetes; or chronic infectious diseases) were excluded. Participants were recruited in-hospital by a study co-ordinator. Approval was obtained from the Research Ethics Committee at each site, and participants provided written informed consent.

Measures

The main dependent variable was leisure-time AEE measured 2, 6 and 12 months after hospitalization. Seven independent variables (sex, age group, education level, reason for hospitalization, CHF, diabetes, and physical activity level before hospitalization) were assessed at recruitment. An eighth variable, participation in cardiac rehabilitation, was assessed at each follow-up time point.

Physical activity

AEE levels after hospitalization were assessed using a telephone-administered 7-day physical activity recall interview [13]. The interviewer obtained estimates of time spent in activities of moderate, hard and very hard intensity [13]. Weekly minute totals at each intensity were converted to hourly values and multiplied by an intensity factor (4, 6 or 10 for moderate, hard, and very hard intensities, respectively) [14] and body weight in kilograms to yield the 7-day AEE (in kcal). The test-retest correlation coefficient for the physical activity recall was 0.63–0.75 and correlations with cardiorespiratory fitness (r = 0.52), accelerometer readings (r = 0.30) and 4-week physical activity history (r = 0.31) were similar to self-report questionnaires [14].

Levels of prehospitalization physical activity were assessed by the leisure score index (LSI) of the Godin Leisure-Time Exercise Questionnaire [15]. The LSI contained three open-ended questions that asked about the average frequency of mild, moderate and strenuous exercise performed in a typical week for at least 15 min each time in the 6-month period before hospitalization. Bouts of moderate and strenuous exercise were multiplied by an intensity factor (5 for moderate and 9 for strenuous) [15] to form a single index of physical activity in metabolic equivalents (of oxygen consumption). LSI scores for combined moderate and strenuous exercise of 24 or greater are consistent with current recommendations for CAD patients (N. Bertrand, G. Godin, 2004, personal communication). The reliability and validity of the LSI compare favorably with other self-report measures of physical activity based on test-retest scores, objective activity monitors, and cardiorespiratory fitness indices [16].

Personal and other variables

Demographic characteristics assessed were sex, age (in years), and education (in years). Age was collapsed to 65 years or less or over 65 years, consistent with categories used in exercise training in CAD patients [17]. Education was categorized as 13 years or less, equivalent to high school graduation or less in Ontario, or more than 13 years, indicating some post-secondary education. The reason for hospitalization was obtained from the hospital chart and categorized as AMI, PCI or CABG. When participants fell into two or more categories (e.g. PCI after AMI, CABG after AMI, CABG after PCI), they were classified on the basis of the event/procedure with the greatest impact on physical activity levels post-hospitalization (CABG>AMI>PCI), determined from a preliminary analysis. Co-morbidities of interest were identified a priori through discussions with clinicians, and included CHF and diabetes. Participants were asked whether a doctor had ever told them they had CHF or diabetes. Participants were asked at each follow-up the following question: ‘Are you currently participating in a structured cardiac rehabilitation program?’ Cardiac patients included in this study were served by three distinct cardiac rehabilitation programs. All programs included core components of cardiac rehabilitation [18], and were of at least 3 months’ duration. Physical activity behaviour was a target for each program and relevant self-management skills were developed in accordance with key constructs of social cognitive theory, including self-efficacy and self-regulation [19]. The standard procedure for the last patient contact before finishing rehabilitation was a conversation with a physiotherapist to update the exercise prescription and to review community resources to support the maintenance of physical activity behaviour after graduation.

Data analyses

Preliminary analyses were conducted to compare study participants with non-participants, explore patterns of missing data, and examine the distribution of data for outliers. A Z-score criterion of 4 or more was used to identify outliers [20]. Descriptive statistics were used to summarize characteristics of the study sample. The trajectory of physical activity behavior was assessed using repeated measures analysis of covariance (ANCOVA), with the within-subject factor being the physical activity level measured at 2, 6 and 12 months and covariates being sex, age group, education level, reason for hospitalization, CHF, diabetes, previous physical activity level, and participation in cardiac rehabilitation. Bonferroni post-hoc tests were used to follow up the main effect for time. To examine the impact of sex, age group, education level, CHF, diabetes, previous physical activity, and cardiac rehabilitation on physical activity over time, a series of two (between-subjects factor) × three (time: physical activity) mixed-model ANCOVA were performed. Regarding the reason for hospitalization, a three (reason) × three (time: physical activity) mixed-model ANCOVA was performed. A Tukey's post-hoc test was used to follow up the main effect for reason for hospitalization. Significant interactions were followed up with a series of repeated measure ANCOVA with time having two levels (i.e. 2 and 6, 6 and 12, or 2 and 12 months).

Results

Recruitment and participant flow

Between May 2002 and December 2003, 1433 patients were approached and 826 (57.6%) agreed to participate. Participants included 308 PCI, 302 AMI, and 216 CABG patients. A survey was completed by 446 non-participants to estimate the representativeness of the sample. There were no differences between participants and non-participants for age, sex, reason for hospitalization, employment status, or location of residence (urban, suburban, or rural). Compared with non-participants, participants were more educated (12.8 versus 11.5 years; P <0.01), less likely to be smokers (16.6 versus 27.9%; P <0.01) and more likely to report a program of regular physical activity before hospitalization (37.5 versus 27.6%; P <0.01).

In the year after the hospitalization at which they were recruited, 25 patients (3.0%) died and were excluded from analyses. An additional 68 patients (8.4%) were re-hospitalized with a new cardiac event (AMI, PCI or CABG), but were retained in the analyses. Of the 801 1-year survivors, complete data were available for 100, 78, 72 and 72% of participants at baseline, 2, 6, and 12 months, respectively. Missing physical activity scores were imputed using the expectation maximization algorithm [21]. Preliminary analyses showed that the distributions of scores for physical activity were heavily skewed. Nineteen outliers were excluded from the analyses, resulting in a normal data distribution. A final sample of 782 was used in analyses.

Characteristics of the study sample

Characteristics of the sample are shown in Table 1. Participants were predominantly male, with a mean age of almost 62 years; 28.1% reported a history of CHF and 20.8% reported diabetes. For the majority (59.8%), this was their first hospitalization for CAD. Before hospitalization, one-quarter of participants reported a physical activity level sufficient to be considered active, using an LSI score of 24 or greater as a criterion. Overall, 31% participated in cardiac rehabilitation.

Trajectory of physical activity levels after hospitalization

There was a significant negative effect of time on physical activity [F(2,773) = 14.7, P<0.001] after adjustment for age, sex, education level, reason for hospitalization, CHF, diabetes, previous level of physical activity, and participation in cardiac rehabilitation.

Table 1

Characteristics of study sample

Variable	Study sample (n = 782)
Age (years) (SD)	61.6 (10.0)
Sex (%)
Men	74.9
Women	25.1
Education (years) (SD)	12.8 (3.6)
Reason for hospitalization (%)
AMI	36.3
PCI	37.1
CABG	26.6
First hospitalization for CAD (%)
Yes	59.8
No	40.2
Activity level before hospitalization (%)
Insufficiently active	74.8
Active	25.2
Participation in cardiac rehabilitation (%)
Did not participate in cardiac rehabilitation	69.2
Did participate in cardiac rehabilitation	30.8

AMI, Acute myocardial infarction; CABG, coronary artery bypass graft surgery; CAD, coronary artery disease; PCI, percutaneous coronary intervention.

Bonferroni post-hocs showed that leisure-time physical activity levels declined significantly between 2 and 6 months [F (1,774) = 28.8; P<0.001] and between 2 and 12 months [F (1,774) = 12.9; P<0.001]; however, there was no difference in weekly physical activity between 6 and 12 months [F (1,774) = 0.5; P = 0.49].

Effect of sex

Summary results are shown in Table 2. The interaction between sex and time was not significant [F (2,773) = 1.8, P = 0.180]. After adjustment for covariates, men reported significantly higher levels of physical activity than women [F (1,774) = 66.4, P = 0.001]. Physical activity levels declined in a similar manner over time in both men and women [F (2,773) = 11.7, P = 0.001].

Effect of age group

Summary results are shown in Table 2. There was a significant interaction (Fig. 1) between age group and time [F (2,773) = 3.2; P = 0.040]. In younger participants, physical activity declined between 2 and 6 months [F (1,478) = 20.7; P<0.001] and there was no significant change between 6 and 12 months [F (1,478) = 0.8; P = 0.368]. In older participants, there was no significant change in physical activity between any of the time points. After adjustment for other covariates, a main effect for age group was found [F (1,774) = 56.7, P <0.001] with younger participants reporting higher levels of physical activity than older participants.

Table 2

Summary data for unadjusted and adjusted repeated measures analysis of covariance for weekly activity energy expenditure by sex, age group, education level, reason for hospitalization, co-morbidities, previous physical activity, and participation in cardiac rehabilitation

Factor	Mean (SD) weekly energy expenditure at moderate intensity or greater			Factor effect F-value	Factor effect adjusted^a F-value	Factor × time F-value	Factor × time adjusted^a F-value
	2 Months	6 Months	12 Months
Sex
Male	2147 (1489)	1834 (1322)	1846 (1565)	78.0∗	66.4∗	1.8	2.1
Female	1356 (1139)	1207 (1060)	1012 (988)
Age group
≤ 65 years	2268 (1544)	1880 (1337)	1862 (1632)	71.8∗	56.7∗	4.4∗∗	3.2∗∗
> 65 years	1440 (1116)	1351 (1140)	1280 (1131)
Education level
≤ High school	1848 (1456)	1509 (1226)	1587 (1543)	14.4∗	1.6	4.9∗∗∗	6.7∗∗∗
Post-secondary	2138 (1423)	1992 (1349)	1731 (1369)
Reason for hospitalization
AMI	1896 (1417)	1651 (1249)	1745 (1480)	3.5∗	3.0∗∗	2.2	1.7
PCI	1926 (1529)	1607 (1357)	1400 (1378)
CABG	2051 (1383)	1805 (1245)	1817 (1599)
Congestive heart failure
No	2112 (1508)	1766 (1305)	1711 (1426)	18.0∗	8.1∗∗∗	3.4∗∗	2.8
Yes	1561 (1225)	1462 (1245)	1479 (1633)
Diabetes
No	1990 (1475)	1750 (1305)	1720 (1486)	9.1∗∗∗	6.3∗∗	1.1	1.1
Yes	1825 (1370)	1424 (1232)	1374 (1483)
Previous level of physical activity
Insufficiently active	1813 (1396)	1553 (1237)	1442 (1400)	47.8∗	30.1∗	2.7	2.1
Active	2348 (1534)	2038 (1374)	2214 (1582)
Cardiac rehabilitation
Did not participate	1850 (1466)	1578 (1288)	1523 (1440)	16.3∗	3.3	0.9	0.1
Did participate	2168 (1391)	1897 (1270)	1892 (1556)

AMI, Acute myocardial infarction; CABG, coronary artery bypass graft surgery; PCI, percutaneous coronary intervention. ^aAdjusted for all other covariates (i.e. sex, age group, education level, reason for hospitalization, congestive heart failure, diabetes, previous physical activity level, and participation in cardiac rehabilitation). ∗P<0.001. ∗∗P<0.05. ∗∗∗P<0.01.

Fig. 1

Age group differences in activity energy expenditure over time after hospitalization for coronary artery disease. ≤ 65 years; <65 years.

Fig. 2

Education level differences in activity energy expenditure over time after hospitalization for coronary artery disease.≤ High school; □ post-secondary.

Effect of education level

Summary results are shown in Table 2. There was a significant interaction (Fig. 2) between education level and time [F (2,773) = 6.7, P = 0.001]. In participants with high school education or less, physical activity declined between 2 and 6 months [F (1,506) = 23.9; P<0.001] although there was no change between 6 and 12 months [F (1,506) = 1.3; P = 0.261]. In participants with at least some post-secondary education, physical activity was lower at 12 months compared with 2 months [F (1,268) = 8.1; P = 0.005]. There was no incremental decline between 2 and 6 months [F (1,268) = 2.6; P = 0.107], but physical activity declined significantly between 6 and 12 months [F (1,268) = 5.3; P = 0.022]. After adjustment for covariates, the main effect of educational level was not significant [F (1,774) = 1.6, P = 0.209].

Fig. 3

Differences in activity energy expenditure over time after hospitalization for coronary artery disease by reason for hospitalization. □ Acute myocardial infarction; □ percutaneous coronary intervention; □ coronary artery bypass graft surgery.

Effect of reason for hospitalization

Summary results are shown in Table 2 and Fig. 3. The interaction between reason for hospitalization and time failed to reach statistical significance [F (4, 1546) = 1.7, P = 0.139]. After adjustment for covariates, a trend for effect of reason for hospitalization on physical activity was observed, which could not be proved significant [F (2,773) = 3.0, P = 0.051]. Because the trend approached significance, Tukey's post-hoc tests were performed. They showed that CABG patients were more physically active over follow-up than PCI patients (P = 0.033). Activity levels in AMI patients were intermediate and were not different from either CABG or PCI patients. Physical activity levels declined over time in all groups [F (2,772) = 10.5, P<0.001].

Effect of congestive heart failure

Summary results are shown in Table 2. The interaction between CHF and time failed to reach statistical significance [F (2,773) = 2.8; P = 0.061]. After adjustment for other covariates, a significant main effect of CHF was observed [F (1,774) = 8.1; P = 0.004]. Participants without CHF were more active than those with a history of CHF. For both those with and without CHF, physical activity declined over time [F (1,774) = 7.6; P<0.001].

Effect of diabetes

Summary results are shown in Table 2. The interaction between diabetes and time was not significant [F (2,773) = 1.1; P = 0.354]. After adjustment for other covariates, a significant main effect of diabetes was observed [F (1,774) = 6.3; P = 0.012]. Participants without diabetes were more active than those with a history of diabetes. For both those with and without diabetes, physical activity declined over time [F (1,774) = 19.3; P<0.001].

Effect of physical activity level before hospitalization

Summary results are shown in Table 2. The interaction between previous physical activity and time after hospitalization failed to reach statistical significance [F (1,774) = 2.1, P = 0.120]. After adjustment for other covariates, participants who reported being physically active before hospitalization had higher levels of physical activity after hospitalization [F (1,774) = 30.1, P <0.001], although physical activity levels declined over time [F (2,773) = 12.3, P <0.001] in both groups.

Effect of participation in cardiac rehabilitation

Summary results are shown in Table 2. The interaction between participation in cardiac rehabilitation and time from hospitalization was not significant [F (2,773) = 0.4, P = 0.956]. After adjustment for other covariates, there was no significant main effect for participation in cardiac rehabilitation [F (1,774) = 3.3; P = 0.072]. Physical activity levels declined over time [F (2,773) = 10.7, P<0.001] for both those who participated in cardiac rehabilitation and those who did not.

Proportion of coronary artery disease patients achieving recommended physical activity levels

For descriptive purposes, the proportions of CAD patients achieving AEE levels of 1000 kcal/week or greater and 1500 kcal/week or greater were calculated (Table 3). Over time, the proportion of participants achieving the 1000 and 1500 kcal/week level declined from 74.6 and 58.6%, respectively, to 61.7 and 45.7%, respectively.

Discussion

This study provides new information on the trajectory of physical activity behavior in the year after hospitalization in a representative sample of patients with CAD. Data showed that physical activity levels in patients participating in and not participating in cardiac rehabilitation decline over time, especially between 2 and 6 months post-hospitalization. Between 2 and 12 months, AEE was reduced by more than 300 kcal/week (∼ 16%), an amount representing approximately 70 min of walking. Age group and education level moderated the trajectory of physical activity behavior; the relative decline was less in older individuals and those with at least some post-secondary education. Independent main effects on physical activity were found for sex (men were more active than women), CHF (those with no history of CHF were more active than those with a history of CHF), diabetes, (those without diabetes were more active than those with diabetes), and previous level of physical activity (patients who were active before hospitalization were more active after hospitalization). A reduction in energy expenditure of the magnitudes described can have an impact on cardiorespiratory fitness, one of the stronger predictors of mortality in CAD patients [22]. For example, we previously found that a 336 kcal/week difference in leisure-time physical activity 6 months after program initiation differentiated participants who improved their cardiorespiratory fitness by 1 or more metabolic equivalent at 1 year from those who did not [4].

Table 3

Proportion of coronary artery disease patients achieving activity energy expenditure levels of 1000 and 1500 kcal/week or more overall and by sex, age group, education level, reason for hospitalization, co-morbidities, previous physical activity, and participation in cardiac rehabilitation

	2 Months		6 Months		12 Months
	≥ 1000 kcal/week	≥ 1500 kcal/week	≥ 1000 kcal/week	≥ 1500 kcal/week	≥ 1000 kcal/week	≥ 1500 kcal/week
All participants	74.6	58.6	65.9	49.5	61.7	45.7
Sex
Male	79.3	65.4	70.2	55.9	68.2	52.7
Female	59.9	37.4	52.4	29.4	41.7	24.1
Age group
≥ 65 years	82.4	69.7	73.9	58.4	65.8	53.2
> 65 years	62.0	40.4	52.7	34.9	55.1	33.6
Education level
≥ High school	71.6	55.1	61.2	41.9	60.0	42.7
Post-secondary	80.1	64.9	74.5	63.5	64.9	51.3
Reason for hospitalization
AMI	74.3	55.8	67.8	48.9	65.2	49.3
PCI	71.6	57.2	60.4	46.0	54.0	40.0
CABG	79.2	64.3	71.0	55.1	67.6	48.8
Congestive heart failure
No	75.9	61.1	68.8	52.7	63.8	47.7
Yes	61.5	43.4	57.9	39.8	54.3	37.6
Diabetes
No	72.5	56.3	68.4	51.6	64.5	47.8
Yes	69.4	55.3	55.9	40.0	48.8	34.1
Previous level of physical activity
Insufficiently active	71.4	53.9	62.8	44.9	55.1	38.6
Active	83.8	72.2	74.7	62.6	80.8	66.2
Cardiac rehabilitation
Did not participate	71.2	55.6	64.0	45.0	58.7	43.6
Did participate	81.9	65.0	70.0	59.3	68.3	50.2

AMI, Acute myocardial infarction; CABG, coronary artery bypass graft surgery; PCI, percutaneous coronary intervention.

Current guidelines indicate that individuals with CAD should expend 1000–1500 kcal/week or more through physical activities. In our study, average weekly energy expenditure exceeded both of these thresholds at each time point; however, there was substantial between-subject variability (SD = 1451 kcal/week with a range of 0–7769 kcal/week). At 1 year, 62 and 46% of participants accumulated 1000 and 1500 kcal/week, respectively.

Activity levels reported in the current study are comparable to other published reports. Schairer et al. [5] studied leisure-time physical activity levels in patients enrolled in a supervised maintenance cardiac rehabilitation program. They found that these patients expended weekly 1504 ± 830 kcal in physical activity, of which 830 ± 428 kcal were expended in cardiac rehabilitation and 675 kcal in leisure-time physical activity; 72% of participants accumulated at least 1000 kcal/week and 43% accumulated 1500 kcal/week or more. Ayabe et al. [23] used an accelerometer to examine physical activity patterns in cardiac rehabilitation program participants. The average weekly energy expenditure was 1597 ± 846 kcal/week and 43% of participants accumulated at least 1500 kcal/week. The proportion of participants in the present study achieving target levels for physical activity was similar to studies including only participants in cardiac rehabilitation. This suggests that many patients with CAD not in cardiac rehabilitation also adopt or continue their own physical activity program.

Our findings concerning the effects of key demographic and previous physical activity levels in patients with CAD are consistent with previous research. There is consistent evidence that men are more active than women during leisure time [5, 11, 24, 25], and that AEE decreases with age [5, 24–26]. Education level is positively associated with the adoption of vigorous physical activity [27], whereas previous physical activity participation predicts long-term physical activity maintenance [12]. Only 42% of women achieved the 1000 kcal/week level at 1-year follow-up compared with more than two-thirds of the men.

This study is the first to report long-term physical activity levels on the basis of the reason for hospitalization. The finding that PCI patients are less physically active than CABG patients is consistent with our clinical experience, and suggests that patients treat PCI as a ‘drive-through’ service, and see little need for a major sustained lifestyle change. PCI patients receive interventions on a short-stay basis, often discharged within 24 h of admission. This leaves less time for education compared with the 4–7 days for AMI and CABG patients, respectively. Given the rapid growth in PCI procedures, there is a need for interventions targeted to this population.

Participation in exercise-based cardiac rehabilitation is associated with lower total and cardiac mortality rates compared with usual medical care [28, 29]. Furthermore, previous reviews of cardiac rehabilitation and secondary prevention have concluded that these programs have long-term effects on physical activity levels [7]. We found that the effects of participation in cardiac rehabilitation on physical activity levels were mostly negated after adjustment for sex, age group, level of education, reason for hospitalization, co-morbidities, and previous physical activity level. This is consistent with previous studies showing biases in referrals to cardiac rehabilitation [30–33], which favor higher levels of physical activity, that is, younger men with higher levels of education referred after CABG who have few co-morbidities and are already physically active. Cardiac rehabilitation programs need new strategies to reach out to patient groups unlikely to attend (e.g. women, previously inactive, PCI patients). For example, Marchionni et al. [34] found home-based programs to be superior to clinic-based programs in the oldest patients. The lack of an observed effect of participation in cardiac rehabilitation on physical activity levels (after adjustment for other covariates) may also be explained by methodological limitations. For example, we have no information regarding the levels of attendance at rehabilitation sessions and the number of program drop-outs. Patients who fail to attend sessions or drop out will not derive as much benefit as those who are adherent. There may have been variability in the professional background and experience of rehabilitation staff, and the quality of program implementation between the three sites that undermined the effects on physical activity.

These results have important implications for those designing activity programs for patients with CAD. Motivation to be active appears to be highest in the early post-discharge period. Health professionals should capitalize on this motivation by providing activity guidelines for patients upon discharge from hospital, rather than waiting until patients attend structured cardiac rehabilitation programs. As the rate of decline in physical activity is greatest between 2 and 6 months and plateaus thereafter, interventions should be intensified during this period. Future studies should assess the mediating effect of interventions provided at strategic time points on the decline in physical activity over time. The effects of sex, age, co-morbidities and previous levels of physical activity appear to be particularly strong. These factors should be used to tailor activity guidance to patients. Tailored activity programs should incorporate varying rates of progression, different levels and types of professional and social support, and be able to accommodate co-morbidities such as CHF and diabetes.

The present study has a number of strengths, including the large sample size, longitudinal follow-up, and the inclusion of large numbers of women, PCI patients, and patients not participating in cardiac rehabilitation. The major limitation is the reliance on self-reported physical activity data; however, there is epidemiological evidence that self-reported physical activity data are quite reliable [35]. In addition, physical activity levels over the past 7 days may not be typical. When generalizing these results to other samples of patients with CAD, readers are reminded that our sample over-represents educated patients and those with a regular program of physical activity before hospitalization. These biases would probably increase reported physical activity levels in our sample.

In summary, physical activity levels in patients who have been hospitalized with CAD decline over time between 2 and 12 months. The trajectory is moderated by age and education level. Sex, co-morbidities and physical activity level before hospitalization, each have an independent effect on physical activity levels after hospitalization. CABG patients are more physically active than PCI patients. These results have implications for targeting interventions to specific subgroups. Further investigation is required to understand the behavioral and environmental mechanisms by which these factors may influence physical activity levels.

References

Leon

Franklin

Costa

Balady

Berra

Stewart

Cardiac rehabilitation and secondary prevention of coronary heart disease: an American Heart Association scientific statement from the Council on Clinical Cardiology (Subcommittee on Exercise, Cardiac Rehabilitation, and Prevention) and the Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity), in collaboration with the American association of Cardiovascular and Pulmonary Rehabilitation. Circulation 2005; 111:369–376.

Gibbons

. In or out - it all adds up. J Cardiopulm Rehabil 2003; 23: 266–268.

Hambrecht

Niebauer

Marburger

Grunze

Kalberer

Hauer

Various intensities of leisure time physical activity in patients with coronary artery disease: effects on cardiorespiratory fitness and progression of coronary atherosclerotic lesions. J Am Coll Cardiol 1993; 22:468–477.

Reid

Dafoe

Morrin

Mayhew

Papadakis

Beaton

Impact of program duration and contact frequency on efficacy and cost of cardiac rehabilitation: results of a randomized trial. Am Heart J 2005; 149:862–868.

Schairer

Keteyian

Ehrman

Brawner

Berkebile

. Leisure time physical activity of patients in maintenance cardiac rehabilitation. J Cardiopulm Rehabil 2003; 23:260–265.

Schairer

Kostelnik

Proffitt

Faitel

Windeler

Rickman

Caloric expenditure during cardiac rehabilitation. J Cardiopulm Rehabil 1998; 18:290–294.

Simons-Morton

Effects of interventions in health care settings on physical activity or cardiorespiratory fitness. Am J Prev Med 1998; 15: 413–430.

Ayala

Orenstein

Greenlund

Croft

Neff

Mensah

. Receipt of cardiac rehabilitation services among heart attack survivors − 19 states and the District of Columbia, 2001. MMWR Morb Mortal Wkly Rep 2003; 52:1072–1075.

Suskin

MacDonald

Swabey

Arthur

Vimr

Tihaliani

. Cardiac rehabilitation and secondary prevention services in Ontario: recommendations from a consensus panel. Can J Cardiol 2003; 19:833–838.

10.

Jolliffe

Rees

Taylor

Thompson

Oldridge

Ebrahim

. Exercise-based rehabilitation for coronary heart disease Cochrane Database Syst Rev 2006; CD001800.

11.

Blanchard

Rodgers

Courneya

Daub

Black

. Self-efficacy and mood in cardiac rehabilitation: should gender be considered? Behav Med 2002; 27:149–160.

12.

Oman

King

. Predicting the adoption and maintenance of exercise participation using self-efficacy and previous exercise participation rates. Am J Health Promot 1998; 12:154–161.

13.

Blair

Haskell

. Assessment of habitual physical activity by seven-day recall in a community survey and controlled experiments. Am J Epidemiol 1985; 122:794–804.

14.

Sallis

. Seven-day physical activity recall Med Sci Sports Exerc 1997; 29 (Suppl 6):S89–S103.

15.

Godin

Jobin

Bouillon

. Assessment of leisure time exercise behavior by self-report: a concurrent validity study. Can J Public Health 1986; 77: 359–362.

16.

Jacobs

Jr Ainsworth

Hartman

Leon

. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc 1993; 25:81–91.

17.

Fletcher

Balady

Amsterdam

Chaitman

Eckel

Fleg

Exercise standards for testing and training: a statement for healthcare professionals from the American Heart Association. Circulation 2001; 104:1694–1740.

18.

Stone

Arthur

, editors. Canadian guidelines for cardiac rehabilitation and cardiovascular disease prevention. 2nd ed. Winnipeg: Canadian Association of Cardiac Rehabilitation; 2004.

19.

Prior

Cupper

. Behavioural, psychosocial and vocational issues in cardiovascular disease. In: Stone

Arthur

, editors. Canadian guidelines for cardiac rehabilitation and cardiovascular disease prevention. Winnipeg: Canadian Association of Cardiac Rehabilitation; 2004.

20.

Tabachnick

Fidell

. Using multivariate statistics, 4th ed. Boston: Allyn and Bacon; 2000.

21.

Cohen

West

. Applied multiple regression/correlation analysis for the behavioral sciences. 3rd ed. Mahwah, NJ: Lawrence Erlbaum Associates; 2003.

22.

Myers

Prakash

Froelicher

Partington

Atwood

. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med 2002; 346:793–801.

23.

Ayabe

Brubaker

Dobrosielski

Miller

Ishi

Yahiro

The physical activity patterns of cardiac rehabilitation program participants. J Cardiopulm Rehabil 2004; 24:80–86.

24.

Bruce

Katzmarzyk

. Canadian population trends in leisure-time physical activity levels, 1981–1998. Can J Appl Physiol 2002; 27:681–690.

25.

Plotnikoff

Mayhew

Birkett

Loucaides

Fodor

. Age, gender, and urban-rural differences in the correlates of physical activity. Prev Med 2004; 39:1115–1125.

26.

King

Castro

Wilcox

Eyler

Sallis

Brownson

. Personal and environmental factors associated with physical inacitivity among different racial-ethnic groups of U.S. middle-aged and older-aged women. Health Psychol 2000; 19:354–364.

27.

Sallis

Hovell

Hofstetter

. Predictors of adoption and maintenance of vigorous physical activity in men and women. Prev Med 1992; 21: 237–251.

28.

Norris

Jensen

Galbraith

Graham

Daub

Knudtson

Ghali

. Referral rate and outcomes of cardiac rehabilitation after cardiac catheterization in a large Canadian city. J Cardiopulm Rehabil 2004; 24:392–400.

29.

Taylor

Brown

Ebrahim

Jolliffe

Noorani

Rees

Exercise-based rehabilitation for patients with coronary heart disease: systematic review and meta-analysis of randomized controlled trials. Am J Med 2004; 116:682–692.

30.

Blackburn

Foody

Sprecher

Park

Apperson-Hansen

Pashkow

. Cardiac rehabilitation participation patterns in a large, tertiary care center: evidence for selection bias. J Cardiopulm Rehabil 2000; 20:189–195.

31.

Bongard

Grenier

Ferrieres

Danchin

Cantet

Amelineau

Cambou

. Drug prescriptions and referral to cardiac rehabilitation after acute coronary events: comparison between men and women in the French PREVENIR Survey. Int J Cardiol 2004; 93:217–223.

32.

Cottin

Cambou

Casillas

Ferrieres

Cantet

Danchin

. Specific profile and referral bias of rehabilitated patients after an acute coronary syndrome. J Cardiopulm Rehabil 2004; 24:38–44.

33.

Williams

Fraser

West

. Gender differences in management after acute myocardial infarction: not ‘sexism’ but a reflection of age at presentation. J Public Health (Oxf) 2004; 26:259–263.

34.

Marchionni

Fattirolli

Fumagalli

Oldridge

Del Lungo

Morosi

Improved exercise tolerance and quality of life with cardiac rehabilitation of older patients after myocardial infarction: results of a randomized, controlled trial. Circulation 2003; 107:2201–2206.

35.

Brownson

Jackson-Thompson

Wilkerson

Kiani

. Reliability of information on chronic disease risk factors collected in the Missouri Behavioral Risk Factor Surveillance System. Epidemiology 1994; 5: 545–549.