The importance of daily physical activity for improved exercise tolerance in heart failure patients with limited access to centre-based cardiac rehabilitation

Abstract

BACKGROUND:

Supervised cardiac rehabilitation provided at dedicated centres ameliorates exercise intolerance in patients with chronic heart failure.

OBJECTIVE:

To correlate the amount of physical activity outside the hospital with improved exercise tolerance in patients with limited access to centre-based programs.

METHODS:

Forty patients (median age 69 years) with stable heart failure due to systolic left ventricular dysfunction participated in cardiac rehabilitation once per week for five months. Using a validated single-axial accelerometer, the number of steps and physical activity-related energy expenditures on nonrehabilitation days were determined.

RESULTS:

Median (interquartile range) peak oxygen consumption was increased from 14.4 mL/kg/min (range 12.9 mL/kg/min to 17.8 mL/kg/min) to 16.4 mL/kg/min (range 13.9 mL/kg/min to 19.1 mL/kg/min); P<0.0001, in association with a decreased slope of the minute ventilation to carbon dioxide production plot (34.2 [range 31.3 to 38.1] versus 32.7 [range 30.3 to 36.5]; P<0.0001). Changes in peak oxygen consumption were correlated with the daily number of steps (P<0.01) and physical activity-related energy expenditures (P<0.05). Furthermore, these changes were significantly correlated with total exercise time per day and time spent for light (≤3 metabolic equivalents) exercise, but not with time spent for moderate/vigorous (>3 metabolic equivalents) exercise.

CONCLUSIONS:

The number of steps and energy expenditures outside the hospital were correlated with improved exercise capacity. An accelerometer may be useful for guiding home-based cardiac rehabilitation.

Cardiac rehabilitation is safe and is associated with both increased exercise capacity and improved symptoms and survival in patients with chronic heart failure (1–6). Therefore, exercise training has been considered to be an important and essential part of heart failure management, and cardiac rehabilitation programs have traditionally been offered at local hospitals with rehabilitation specialists and facilities. Centre-based programs are generally resource-intensive and are limited by accessibility and a low long-term adherence rate, especially for elderly patients with heart failure (7,8). To overcome these limitations, home-based cardiac rehabilitation is widely used for patients who may have difficulties participating in centre-based programs.

Whether home-based cardiac rehabilitation is equally effective in the treatment of heart failure is controversial (7,9,10). In this context, a relatively small amount of improvement in maximal exercise capacity observed in the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial may be attributed, at least in part, to low-dose exercise during a nonsupervised rehabilitation phase (11). However, data on actual exercise intensities in home-based programs or its relationship to the degree of improvement in exercise capacity are lacking. Thus, we hypothesized that the magnitude of increased exercise capacity is related to daily physical activities outside the hospital. To assess physical activities in a free-living condition, we used an accelerometer and aimed to correlate the daily number of steps and physical activity-related energy expenditures with changes in peak oxygen consumption (VO₂).

METHODS

Patients

Forty patients (34 men and six women) with chronic heart failure caused by left ventricular systolic dysfunction were included in the present study. These patients could not fully participate in the standard cardiac rehabilitation program (three times per week) because of limited accessibility to the hospital, and were only able to attend once per week. Clinical characteristics of the participants are summarized in Table 1. All patients were symptomatic, with left ventricular ejection fractions of 45% or less, and had been in a clinically stable condition for at least one month. Exclusion criteria were as follows: exercise-induced myocardial ischemia, chronic restrictive or obstructive pulmonary disease, stroke, peripheral artery disease and other noncardiac exercise-limiting disorders. The present study was conducted in accordance with hospital guidelines, and all patients provided written informed consent before the study.

TABLE 1.

Sample baseline characteristics (n=40)

Age, years	69 (62 to 75)
Sex
Male	34 (85)
Female	6 (15)
Weight, kg	60.9 (56.6 to 66.8)
Body mass index, kg/m2	22.8 (21.2 to 24.8)
Etiology of heart failure (ischemic)	21 (53)
NYHA functional class
II	17 (42)
III	23 (58)
Left ventricular ejection fraction, %	34.1 (25.8 to 43.8)
BNP, pg/mL	213.9 (91.0 to 424.5)
Peak oxygen consumption, mL/kg/min	14.4 (12.9 to 17.8)
VE/VCO2 slope	34.2 (31.3 to 38.1)
Medication
Digitalis	11 (28)
ACEI and/or ARB	35 (89)
Beta-blocker	31 (78)
Calcium channel blocker	7 (18)
Statin	25 (63)

Data are presented as median (interquartile range) or n (%). ACEI Angiotensin-converting enzyme inhibitor; ARB Angiotensin-receptor blocker; BNP Brain-type natriuretic peptide; NYHA New York Heart Association; VCO₂ Carbon dioxide production; VE Minute ventilation

Study protocol and exercise training

The cardiac rehabilitation program included aerobic exercise training and educational programs, and has been described elsewhere (12). Briefly, each 90 min session consisted of a warm-up of light exercise, bicycle ergometer and walking on an indoor 60 m track, and cool-down exercise. During the training, electrocardiograms were continuously telemonitored. The intensity of exercise was prescribed for each patient according to the results of cardiopulmonary exercise testing; target heart rate and work load were prescribed at the point of ventilatory anaerobic threshold and at 1 min before the anaerobic threshold, respectively. Median (interquartile range) exercise intensity that was prescribed for the studied participants was 72.4% (range 64.6% to 78.6%) of peak oxygen consumption (VO₂) or 3.2 metabolic equivalents (METs; 1 MET=3.5 mL/kg/min; range 2.9 METs to 3.6 METs). With regard to exercise training outside the hospital, the patients were instructed to walk for at least 20 min per day at a rate of perceived exertion of 13 on the modified Borg scale (13). Clinical evaluation was performed at the beginning of the study and after five months.

Accelerometer

For measurement of the daily number of steps and energy expenditure, the patients wore a single-axial accelerometer (Lifecorder, Suzuken Co, Japan) on a belt at waist level continuously for seven days, except while bathing and sleeping. Data were downloaded to a personal computer and the daily steps and energy expenditures taken on non-rehabilitation days were assessed. The measurements were performed twice during a five-month period of cardiac rehabilitation.

Specific features of the Lifecorder accelerometer have been reported previously (8,14). The accelerometer used in the present study was small (62.5 mm × 46.5 mm × 26.0 mm) and lightweight (40 g). The Lifecorder accelerometer samples vertical acceleration that ranges from 0.06 g to 1.94 g (1 g = acceleration due to Earth’s gravity) and determines an intensity of physical activity every 4 s. Previous validation studies have demonstrated that total energy expenditures measured by the device were highly correlated with those measured by other methods, including the doubly labelled water method, respiratory chamber method and daily record method (14–16). Additionally, the device is able to determine movement intensities on arbitrary scales between 1 (minimal intensity) and 9 (maximal intensity) every 4 s, and the determined intensity level has been demonstrated to correlate closely with METs (14). In the present analysis, intensity levels were arbitratily defined as light (intensity level 1 to 3, or 1.8 to 3 METs), moderate (intensity level 4 to 6, or 3 to 6 METs), or vigorous (intensity level 7 to 9, or ≥6 METs). The time spent for each activity level was calculated separately.

Cardiopulmonary exercise testing

All participants underwent cardiopulmonary exercise testing at the beginning and at the end of the study. A ramp protocol was selected according to the subjective assessment of the patient’s functional capacity to ensure appropriate exercise duration (approximately 10 min to 15 min) and avoid premature fatigue (12). Breath-by-breath gas exchange analysis was performed using a dedicated metabolic cart (Model AE300S, Minato Medical Science, Japan). Expiratory gas was continuously analyzed for the determination of minute ventilation (VE), VO₂ and carbon dioxide production (VCO₂). Peak VO₂ was defined as the highest VO₂ achieved by the patient. Ventilatory response to exercise was assessed using the VE/VCO₂ slope, which was derived from a simple regression analysis of the VE versus VCO₂ plot. Measurements were based on a single test.

Statistics

Data are presented as median (interquartile range). Comparisons of continuous variables before and after cardiac rehabilitation were performed using Wilcoxon’s signed rank test. A linear regression analysis was used for correlations of two continuous variables. Differences were considered to be statistically significant at P<0.05.

RESULTS

All 40 patients completed the five-month, once-weekly cardiac rehabilitation program without worsening heart failure. Body mass index did not change (22.8 kg/m² [range 21.2 kg/m² to 24.8 kg/m²] versus 22.9 kg/m² [range 21.4 kg/m² to 25.0 kg/m²]) over the study period. Plasma brain-type natriuretic peptide concentrations were decreased significantly from 213.9 pg/mL (range 91.0 pg/mL to 424.5 pg/mL) to 89.9 pg/mL (range 35.4 pg/mL to 219.1 pg/mL) after five months (P<0.0001).

Maximal exercise capacity

Peak VO₂ was significantly increased by 9.8% (range 0.8% to 17.6%) from 14.4 mL/kg/min (range 12.9 mL/kg/min to 17.8 mL/kg/min) to 16.4 mL/kg/min (range 13.9 mL/kg/min to 19.1 mL/kg/min) (P<0.0001). As indicated in Figure 1, there was a statistically significant positive correlation between changes in peak VO₂, (reported as % increase in peak VO₂) and mean number of daily steps (P<0.001, r²=0.291). There was also a positive correlation between % increases in peak VO₂ and mean daily physical activity-related energy expenditures for both absolute and body weight-adjusted values (P<0.01, Figure 2).

Figure 1).

Correlation between the daily number of steps and % changes in peak oxygen consumption (%ΔPVO₂)

Figure 2).

Correlations between daily physical activity-related energy expenditures (PAEE) and % changes in peak oxygen consumption (% ΔPVO₂). PAEE are reported as absolute values in kcal/day (A) and as body weight (BW)-adjusted values in kcal/kg/day (B)

The mean time spent for light exercise (≤3 METs) and that for moderate/vigorous exercise (>3 METs) on nonrehabilitation days were 46.0 min/day (range 31.0 min/day to 62.0 min/day) and 8.0 min/day (range 2.4 min/day to 14.5 min/day), respectively. Increases in peak VO₂ were associated with the total exercise time per day (P<0.001, Figure 3A) and with the time spent for light exercise (P<0.001, Figure 3B), but not with the time spent for moderate/vigorous exercise (Figure 3C).

Figure 3).

Per cent changes in peak oxygen consumption (%ΔPVO₂) are plotted against the total daily exercise time (A), the time spent for light exercise (B) and the time spent for moderate to vigorous exercise (C). n.s Not statistically significant

Ventilatory response to exercise

VE/VCO₂ slope was significantly decreased after five months from 34.2 (range 31.3 to 38.1) to 32.7 (range 30.3 to 36.5) (P<0.0001). However, changes in the VE/VCO₂ slope did not correlate with the number of steps or physical activity-related energy expenditures (Figure 4).

Figure 4).

Per cent changes in minute ventilation/carbon dioxide production slope (%ΔVE/VCO₂) are plotted against the daily number of steps (A) and body weight (BW)-adjusted physical activity-related energy expenditures (PAEE) (B). n.s Not statistically significant

DISCUSSION

We studied relatively elderly patients (median age 69 years) with functional class II or III heart failure who were unable to fully participate in a centre-based program (three or more sessions per week), and demonstrated that there was a positive correlation between changes in peak VO₂ and the extent of free-living physical activities, as assessed by the daily number of steps and estimated physical activity-related energy expenditure. These results support the hypothesis that physical activity outside of the hospital is an important determinant for improved exercise tolerance.

In the present study, the increase in peak VO₂ after five months was 9.8%, a smaller increase than previously reported (11,17). We prescribed exercise dose according to the results of cardiopulmonary exercise testing of each patient (72.4 % of peak VO₂ [range 64.6% of peak VO₂ to 78.6% of peak VO₂] or 3.2 METs [range 2.9 METs to 3.6 METs]) and encouraged the patient to exercise for at least 20 min per day. However, the actual median time spent for activities of 3 METs or more was only 8 min. Such low exercise intensities at home and a relatively short study duration (five months) may be the reason for small increases in peak VO₂ in these patients. In addition, these patients were somewhat older compared with those enrolled in clinical trials. Elderly patients are known to experience chronotropic incompetence, increased arterial stiffness and low vital capacity, all of which may limit the beneficial effects of physical training on exercise tolerance (18).

Intensity of physical activities

The optimal intensity or frequency of exercise training for heart failure patients has not been established. The American College of Cardiology/American Heart Association 2005 Guidelines for the management of heart failure patients recognize exercise training as an important therapeutic approach, but do not specify exercise dose (19). Exercise at 60% to 80% of maximal VO₂ for 20 min to 60 min, three to five times per week was adopted in the majority of earlier studies (13). However, this amount of exercise appears to be excessive for elderly patients with heart failure and may lower the adherence rate, especially at home. For example, Ayabe et al (8) reported that physical activity was lower on nonrehabilitation days than on centre-based rehabilitation days. In this context, our observation that there was a positive correlation between the time spent for light exercise (≤3 METs) and improved peak VO₂ may be of therapeutic importance in the management of elderly heart failure patients. Many elderly patients have concomitant exercise-limiting morbidities, such as neuromuscular or orthopedic problems, and it may be appropriate to encourage such patients to exercise at a lower intensity than has been considered necessary to increase maximal exercise capacity. However, we do not disregard the beneficial effects of moderate/vigorous exercise in cardiac rehabilitation programs. The time spent for >3 METs exercise was short in our patients and we were unable to further analyze the effect of such exercise.

Accelerometer

A single-axis accelerometer was used for the quantification of energy expenditures related to physical activities. The accuracy of this device in determining physical activity-related energy expenditures was tested in different cohorts, including elderly subjects. Although the correlation between accelerometer-derived energy expenditures and those determined by other modalities, such as the doubly labelled water method, was satisfactorily high, the accelerometer underestimated energy expenditures by 8% to 20% (14,15,20,21). It should be noted, however, that the primary aim of the present study was not to measure true energy expenditures but to correlate the amount of daily activities to changes in exercise capacity. Intriguingly, the number of daily steps were also correlated with changes in peak VO₂ and may be of practical usefulness for both patients and cardiac rehabilitation professionals.

Limitations

First, the number of enrolled patients in the present study was relatively small, and it was not possible to evaluate the relationship in subgroups divided according to age, etiology of heart failure or baseline exercise capacity. For the same reason, we could not determine factors other than daily acitivity that should have contributed to improved exercise tolerance. Our observation needs to be tested in a larger number of patients and a simple pedometer may be useful for this purpose. Second, the correlation between physical activities and improved exercise tolerance does not necessarily indicate a unidirectional, causal relationship. A higher physical activity level may result in a greater exercise capacity which, in turn, leads to more daily physical activities. The relationship may thus be mutual and there should be positive feedback between physical activities and exercise tolerance.

Conclusions

The number of steps and physical activity-related energy expenditure as well as the time spent for relatively light exercise correlated with the degree of increase in peak VO₂ in patients with chronic heart failure and limited access to centre-based cardiac rehabilitation. A uniaxial accelerometer (or a simple pedometer) may be useful for guiding cardiac rehabilitation participants and should be incorporated into home-based programs.