The purpose of this study was to determine both an association between mortality and physical activity (PA) objectively measured by accelerometer and cutoff values for PA in Japanese outpatients with heart failure (HF). This prospective observational study comprised 170 HF outpatients (mean age, 65.2 years; 77% men). Peak oxygen uptake (VO 2 ) and the relation between ventilation and carbon dioxide production (VE/VCO 2 slope) as indices of exercise capacity were measured during cardiopulmonary exercise testing with a cycle ergometer. PA was assessed by accelerometer-measured average step count (steps) per day for 1 week. Study endpoint was cardiovascular-related death. Over an average follow-up of 1,377.1 (median, 1,335) days, 31 cardiovascular-related deaths occurred. Patients were then divided into survivor (n = 139) and nonsurvivor (n = 31) groups. Brain natriuretic peptide level was significantly different between groups. Peak VO 2 and steps were also significantly lower and VE/VCO 2 slope higher in the nonsurvivors versus survivors. Univariate Cox proportional hazards analysis showed brain natriuretic peptide, peak VO 2 , VE/VCO 2 slope, and steps to be significant prognostic indicators of survival. Multivariate analysis showed PA of ≤4,889.4 steps/day to be a strong and independent predictor of prognosis (hazard ratio: 2.28, 95% confidence interval: 1.31–6.30; p = 0.008). Kaplan-Meier curves after log-rank test showed significant prognostic difference between PA of ≤4,889.4 and >4,889.4 steps/day in the 2 groups (log-rank: 12.19; p = 0.0005). In conclusion, step count as objectively measured by accelerometer may be a prognostic indicator of mortality in Japanese outpatients with HF.
Little is known about the relation between cardiovascular-related mortality and physical activity (PA) as objectively measured by step count with an accelerometer and other survival-related parameters such as left ventricular ejection fraction (LVEF), plasma brain natriuretic peptide (BNP) concentration, peak oxygen uptake (peak VO 2 ), and VE/VCO 2 slope (relation between ventilation [VE] and carbon dioxide production [VCO 2 ]) in heart failure (HF) outpatients. We hypothesized that low PA, as measured by step count, would be related to a reduced rate of survival and might predict mortality in addition to other survival-related parameters in HF outpatients. The present study aimed to determine (1) the relation among peak VO 2 , VE/VCO 2 slope, PA, and mortality by investigating whether these values are related to differences in long-term cardiac-related mortality, and (2) whether there are cut-off values for peak VO 2 , VE/VCO 2 slope, and PA as measured by step count that are associated with a reduction in mortality between survivors and nonsurvivors in a cohort of Japanese outpatients with HF.
Methods
In this prospective observational study, consecutive Japanese outpatients with HF were selected for evaluation by cardiopulmonary exercise testing from 1,899 cardiac patients with myocardial infarction, coronary artery bypass grafting, valve replacement, or HF who visited St. Marianna University School of Medicine Hospital from November 2002 to October 2010. Patients classified as New York Heart Association functional class IV and those with neurological, peripheral vascular, orthopedic, or pulmonary disease were excluded. From these 1,899 patients, 477 stable HF outpatients aged 50 to 89 years were included in this study, which was approved by the St. Marianna University School of Medicine Institutional Committee on Human Research. Written informed consent was obtained from each patient.
Clinical characteristics were evaluated by review of medical records and included age, gender, body mass index, LVEF, etiology of HF, and medications. A cardiologist assessed LVEF as the index of cardiac function by echocardiography in all patients. Cardiopulmonary exercise testing was performed with a cycle ergometer ramp protocol as previously described and only on patients who had not undergone any exercise testing previously. Testing included an initial 3-minute rest on the cycle ergometer followed by 3 minutes of warm-up at a 10- or 20-watt load and full exercise with a linear increase in load by 1 watt every 6 seconds. A 12-lead electrocardiogram from which heart rate was determined was continuously monitored throughout the test. The patient’s VO 2 , VE, and VCO 2 were measured breath by breath throughout the exercise period from expired gases with an AE-300S aero monitor (Minato Ikagaku Co., Tokyo, Japan) and calculated on a personal computer. The exercise test endpoint as determined by the criteria of the American College of Sports Medicine included leg fatigue, shortness of breath, or respiratory exchange ratio >1.15. Peak VO 2 and VE/VCO 2 slope were calculated as indices of exercise capacity.
We used average daily number of steps taken per day over 7 days as the index of PA. This index was measured with a Kenz Lifecorder uniaxial accelerometer (Suzuken Co., Ltd., Nagoya, Japan), which provides reliable and validated output data. Step count is recorded after age, gender, height, and weight data of the individual patient are entered. All patients wore the accelerometer on a belt at waist level just above either leg for 24 hours a day for 8 days except while bathing and sleeping. After the eighth day, the device was retrieved, and the data were downloaded onto a computer and analyzed with Microsoft Excel software. To assess each patient’s usual daily PA, we used the final 7 days (1 week) of continuous data from the 8-day collection period. The average daily number of steps taken over 1 week was calculated as follows: total step count over 7 days/7. Outcome of the study was mortality from cardiac-related death and was evaluated between August 2012 and September 2012. Survival or nonsurvival was determined through review of each patient’s medical records.
Results are expressed as mean ± SD. Parametric and chi-squared tests were used to analyze differences between the survivor and nonsurvivor groups. Because comparisons between these 2 groups were performed for the HF sample across the clinical characteristics of the patients, cardiopulmonary exercise test values, and PA, the Mann-Whitney U test was used to test for differences between 2 independent groups. Receiver operating characteristic (ROC) curves were constructed by plotting true-positive rates (sensitivity) against false-positive rates (1-specificity) to determine the best cut-off value for differences in these values between the survivor and nonsurvivor groups. The area under the curve (AUC) and 95% confidence interval (CI) were also calculated. AUC values >0.9 indicate high accuracy, 0.7 to 0.9 indicate moderate accuracy, and 0.5 to 0.7 indicate low accuracy. After calculating these values with the ROC curves, cut-off values for predicting mortality were chosen. If significant values for clinical characteristics, cardiopulmonary exercise testing, and PA were noted, we identified these cut-off values for calculation using univariate and multivariate Cox proportional hazards models. A univariate Cox proportional hazards model was used to measure the effect of clinical characteristics, exercise capacity, and PA on survival time. Multivariate Cox proportional hazards analysis was also performed for the subjects who had variables entered in the univariate analysis. Differences in survival and nonsurvival were detected by Kaplan-Meier survival curves and compared using the log-rank test. A p value of <0.05 was considered to indicate statistical significance. Statistical analyses were performed with IBM SPSS 17.0 J statistical software (IBM SPSS Japan, Inc., Tokyo, Japan).
Results
Of the 477 potential subjects, 201 met the inclusion criteria, and 276 subjects were excluded because they had undergone previous cardiopulmonary exercise testing. Of these 201 patients, 27 additional patients were excluded because of insufficient data to evaluate their patient clinical characteristics, cardiopulmonary exercise test data, and/or PA. Thus, 174 patients were included for analysis on the basis of the inclusion criteria. Over a mean follow-up period of 1,377.1 ± 747.9 days, 35 of the 174 patients died due to sudden cardiac death (n = 10), progressive HF (n = 21), and non-cardiovascular-related disease (traffic accident, cancer in 2 patients, and gastrointestinal hemorrhage; n = 4). After excluding these 4 patients from the study, we compared patient clinical characteristics and mortality between 139 survivors and 31 nonsurvivors.
Clinical characteristics and mortality data of the study patients are presented in Table 1 . Patient characteristics were similar between the 2 groups with the exception of BNP level, which was significantly higher in the nonsurvivors versus survivors. ROC curve analysis indicated the cut-off value for BNP to be 201.8 pg/dl (sensitivity, 67%; specificity, 35%; AUC = 0.697; 95% CI: 0.608–0.787; p <0.001). No patient experienced chest pain or ischemic ST changes or serious arrhythmia during exercise testing. Respiratory exchange ratio as one endpoint of exercise testing was not significantly different between survivors and nonsurvivors (1.2 ± 0.6 vs 1.2 ± 0.8; p = 0.261). Mortality in relation to peak VO 2 , VE/VCO 2 slope, and PA in the HF outpatients is shown in Table 2 . Peak VO 2 was significantly lower, and VE/VCO 2 slope significantly higher, in the nonsurvivors than the survivors. The cut-off value for peak VO 2 by ROC curve analysis was 15.6 ml/min/kg (sensitivity, 72%; specificity, 49%; AUC = 0.711; 95% CI: 0.621–0.798; p <0.001), and that for VE/VCO 2 slope was 33.7 (sensitivity, 73%; specificity, 45%; AUC = 0.671; 95% CI: 0.580–0.762; p <0.001). Step count (PA) was significantly lower in the nonsurvivors than survivors. ROC curve analysis identified a cut-off value of 4,889.4 steps/day as the optimal predictor of mortality, with a sensitivity of 72% and 1-specificity of 37% (AUC = 0.742; 95% CI: 0.671–0.800; p <0.001) for the prediction of survival time ( Figure 1 ).
| Variable | Total | Survivors | Nonsurvivors | T or χ 2 Value | p Value |
|---|---|---|---|---|---|
| (n = 170) | (n = 139) | (n = 31) | |||
| Age (yrs) | 65.2 ± 8.5 | 64.6 ± 8.3 | 67.6 ± 9.1 | 1.82 | 0.08 |
| Male gender | 77% | 76% | 89% | 2.51 ∗ | 0.16 |
| BMI (kg/m 2 ) | 22.6 ± 3.6 | 22.6 ± 3.1 | 22.4 ± 5.1 | −0.31 | 0.75 |
| LVEF (%) | 35 ± 11 | 35 ± 11 | 32 ± 10 | −1.91 | 0.08 |
| BNP (pg/mL) | 244.9 ± 252.9 | 224.1 ± 260.5 | 324.9 ± 205.7 | 2.14 | 0.02 |
| Etiology | |||||
| Dilated cardiomyopathy | 50% | 50% | 49% | 2.24 ∗ | 0.52 |
| Previous MI | 26% | 24% | 34% | ||
| Arrhythmia | 16% | 17% | 12% | ||
| Medications | |||||
| Beta blockers | 85% | 86% | 80% | 0.72 ∗ | 0.43 |
| ARB | 45% | 46% | 42% | 0.10 ∗ | 0.84 |
| ACEI | 44% | 44% | 45% | 0.02 ∗ | 1.00 |
| Diuretic | 88% | 87% | 94% | 1.36 ∗ | 0.37 |
| Total | Survivors | Nonsurvivors | T Value | p Value | |
|---|---|---|---|---|---|
| No. of patients | 170 | 139 | 31 | ||
| Peak oxygen uptake | 18.2 ± 5.2 | 18.9 ± 5.3 | 15.3 ± 3.2 | −5.11 | <0.001 |
| VE/VCO 2 slope | 34.8 ± 8.0 | 33.9 ± 6.9 | 38.8 ± 10.6 | 3.27 | <0.001 |
| No. of steps | 5,987.3 ± 2,741.8 | 6,411.6 ± 2,793.1 | 4,350.9 ± 1,774.1 | −5.36 | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
