No effect of group-based aerobic interval training on N-terminal pro- B-type natriuretic peptide levels in patients with chronic heart failure

Abstract

Objectives. Exercise training might improve cardiac function as well as functional capacity in patients with chronic heart failure (CHF). N-terminal pro-B-type natriuretic peptide (NT pro-BNP), is associated with the severity of the disease, and has been reported to be an independent predictor of outcome in CHF. We evaluated the effect of a four months group-based aerobic interval training program on circulating levels of NT pro-BNP in patients with CHF. We have previously reported improved functional capacity in 80 patients after exercise in this exercise program. Methods. Seventy-eight patients with stable CHF (21% women; 70±8 years; left ventricular ejection fraction 30±8.6%) on optimal medical treatment were randomized either to interval training (n=39), or to a control group (n=39). Circulating levels of NT pro-BNP, a six minute walk test (6MWT) and cycle ergometer test were evaluated at baseline, post exercise, and further after 12 months. Results. There were no significant differences in NT pro-BNP levels from baseline to either post exercise or long-term follow-up between or within the groups. Inverse correlations were observed between NT pro-BNP and 6MWT (r=−0.24, p=0.035) and cycle exercise time (r=−0.48, p<0.001) at baseline. But no significant correlations were observed between change in NT pro-BNP and change in functional capacity (6MWT; r=0.12, p=0.33, cycle exercise time; r=0.04, p=0.72). Conclusion. No significant changes in NT pro-BNP levels were observed after interval training, despite significant improvement of functional capacity.

Key words::

• Heart failure
• NT pro-BNP
• exercise training
• physical therapy

Exercise training is a valid supplementary treatment in patients with chronic heart failure (CHF). Controlled clinical trials have demonstrated that exercise training programs of various intensity improve functional capacity, quality of life and New York Heart Association (NYHA) class (1), in patients with CHF. Two meta-analyses also indicate that there is a significant reduction in the combined end point of death or all-cause hospitalization (2,3). In the largest intervention trial of exercise treatment in CHF, the HF-ACTION trial, a modest but statistically not significant reduction in all-cause mortality or hospitalization was found (4). Neurohormonal factors such as plasma levels of brain natriuretic peptide (BNP) or the aminoterminal fragment N-terminal pro-B-type natriuretic peptide (NT pro-BNP) of its prohormone (proBNP) have shown to be useful markers of severity in patients with CHF (5), and a reduction in NT pro-BNP levels has been suggested as a significant marker for improved prognosis (6). The six minute walk test (6MWT) is also related to severity (7) and prognosis (8) in patients with CHF. Some studies have suggested that exercise training offer additional non-pharmacological reduction in circulating levels of NT pro-BNP in patients with CHF (9), but there are conflicting results (10,11).

Most exercise training programs for patients with CHF evaluate continuous aerobic exercise programs on treadmills or cycle. In the present study, we investigated the effect of a group-based high-intensity interval-training program on circulating levels of NT pro-BNP in patients with stable CHF. Additionally, the relationship between NT pro-BNP and functional capacity was evaluated.

Materials and methods

Study population

All patients in the present study were part of a CHF study conducted at Oslo University Hospital, Ullevaal, Norway. The design, inclusion, and exclusion criteria have been described in details elsewhere (12). In brief, 78 patients with a mean age 70±8 years, stable CHF and NYHA class II–IIIB were recruited from the heart failure outpatient clinic. All patients referred to the heart failure clinic in a 2.5 years period were screened for inclusion. Criteria for inclusion were symptoms and/or signs of heart failure and a reduced left ventricular ejection fraction (EF) (≤40%) or preserved left ventricular EF (>40%). All patients were in stable condition for at last four weeks before entering into the study, and they were all on optimal medical therapy (Table I). At baseline, they were randomly assigned to a four months group-based aerobic interval training program (n=39) or to a control group (n=39), and followed prospectively for 12 months.

Table I. Baseline characteristics of the study population. Data are presented as mean value±SD or proportions.

Characteristic	Exercise group (n = 39)	Control group (n = 39)	p value
Mean age (years)	68.9±7.9	71.5±7.9	0.14
Gender (male/ female)	30/9	31/8	1.0
NYHA class II/III	21/18	24/15	0.36
LVEF %	29.9±7.6	30.9±9.6	0.62
LVEF>40%	4	3	0.44
Heart rate	67±5	71±15	0.25
Systolic blood pressure (mmHg)	124±13	128±30	0.54
Diastolic blood pressure (mmHg)	79±11	74± 14	0.67
Creatinine (μmol/l)	104±42	99±27	0.50
GFR (ml/min/1.73m^2)	64 ± 17	65 ± 15	0.93
Glucose (mmol/L)	5.8± 1.2	6.1 ±1.9	0.34
Diabetes	9 (23%)	5 (13%)	0.24
Pacemaker/ICD	3 (8%)	4 (10%)	1.0
COPD	6 (15%)	4 (10%)	0.50
Atrial fibrillation	17 (43.5)	11 (28%)	0.16
Primary etiology
IHD	27 (69%)	27 (69%)	1.0
Hypertension	5 (13%)	5 (13%)	1.0
Cardiomyopathy	7 (18%)	7 (18%)	1.0
Drugs
ACE inhibitors/AT-II blockers	38 (97.5%)	39 (100%)	1.0
β-blockers	37 (95%)	35 (90%)	0.68
Loop diuretic	37 (95%)	36 (92%)	0.64
Digitalis	9 (23%)	13 (33%)	0.31

NYHA=New York Heart Association, LVEF=Left ventricular ejection fraction, GFR=Glomerular filtration rate, COPD = chronic obstructive pulmonary disease, IHD = Ischemic heart disease, ACE=Angiotensin converting enzyme, AT-II=Angiotensin receptor II blocker.

p values refer to differences between groups.

Exercise program

A detailed description of the exercise program has been previously reported (Norwegian Ullevaal model) (13). In brief, the program consisted of group-based simple aerobic exercises, including three intervals of high intensity (15–18 on the Borg scale). The total duration of the exercise program was 50 minutes. In addition, the exercise program was followed by 15–30 minutes of group-based counseling after each exercise session, and four individual lessons of counseling by a CHF nurse. The program was offered twice a week for four months. The control group (n=39) was referred to follow-up care by their primary care physician and was not discouraged from regular physical activity.

Measurements

Patients were evaluated at baseline (T1), after 32 sessions of exercise training (within 24 weeks) (T2), and after 12 months (T3). All tests were performed on full medication. Blood samples were obtained at rest between 8 am and 9 am, in a sitting position, after an overnight fast. Levels of NT pro-BNP were determined in EDTA-plasma with Elecsys proBNP sandwich immunoassay on Elecsys 2010 (Roche Diagnotics, Indianapolis, IN, USA). The inter-assay coefficient of variation in our laboratory was 7%. GFR was calculated according to the MDRD formula (14). Baseline tests included 6MWT, a symptom limited exercise test on an electrically braked cycle ergometer (Ergometrics 800, Ergoline, Bitz, Germany) with a stepwise increased workload by 10 W every minute (starting at 30 W), and a Minnesota Living with Heart Failure Questionnaire (MLHFQ).

The study was approved by the Regional Committee for Medical Research Ethics and the Norwegian Data Inspectorate. All patients gave their written informed consent.

Statistical analysis

Sample size calculation was based on the change in the primary outcome measurement for this randomized controlled trial, the 6MWT. Sample size or power calculation for this sub study on NT pro-BNP, was not performed prior to the start of the study. Independent-samples Student's t-test was performed to compare the groups for continuous variables and with χ² square test for categorical data. Because of skewed distribution group comparisons of the NT pro-BNP levels were conducted using the Mann Whitney test. Analyses of intra-group changes in NT pro-BNP were performed using Wilcoxon signed rank test. Differences in changes from baseline to T2 and from baseline to T3, between the randomized groups, were assessed by an ANCOVA model, with the baseline NT pro-BNP levels as a covariate (skewed data were log-transformed before entered into the regression model). Correlation analyses between the variables were conducted with the Spearman's rho. Blood samples were conducted by per protocol analyses and functional capacity and quality of life by intent-to-treat analyses using the last-observation-carried-forward principle.

All statistical analyses were performed using the software package SPSS, version 15.0 (SPSS Inc., Chicago, Illinois, USA). Data are expressed as mean ± SD unless otherwise stated, and p-value < 0.05 was considered statistically significant.

Results

The exercise program was well tolerated, with no complications during exercise. After randomization, four patients discontinued the study. One patient died (control group); one dropped out because of dissatisfaction with group allocation (control group); one dropped out because a cerebral stroke a short period after randomization, but before start of exercising (exercise group); and one did not give reason for drop-out (exercise group). Of the 74 patients who completed the T2, 70 returned for evaluation one year after time 1. In the exercise group, two died and one refused to participate in the follow-up tests. In the control group, one refused to participate in the follow-up tests. Baseline characteristics of the total study population according to the randomized groups are listed in Table I. There were no significant differences between the two groups at baseline, except for the MLHFQ (Table II).

Table II. NT pro-BNP levels, functional capacity and Quality of life at baseline (T1), post exercise (T2) and long-term follow-up (T3) according to the randomized groups. Data are presented as mean value±SD if not stated otherwise.

	Control group (n=39)			Exercise group (n=39)
Variable	Baseline (T1)	Post exercise (T2)	Follow-up (T3)	Baseline (T1)	Post exercise (T2)	Follow-up (T3)	p value^2 T1-T2	p value^2 T1-T3
NT pro-BNP	235 (131, 392)	165 (78, 302)	166 (91, 287)	167 (89, 294)	198 (92, 375)	158 (44, 256)	0.22	0.38
(pmol/l)^1	(n=39)	(n=37)	(n=35)	(n=39)	(n=33)	(n=34)
6MWT (m)	454±84	438±101	433±118	456±76	511±92	496±100	<0.001	<0.001
Cycle exercise time (sec)	337+97	327±120	328±122	345±111	402±156	397±155	<0.001	0.003
Workload (watt)	73±17	72±19	71±21	75±18	84±26	84±26	<0.001	0.001
MLHFQ score	22±17	23±20	28±20	33±18*	22±12	22±14	0.034	0.002

6MWT = six minute walk test, MLHFQ = Minnesota Living with Heart failure Questionnaire (higher scores indicating a lower quality of life).

¹medians (25, 75 percentiles)

²p values refer to difference in change from baseline to T2 and from baseline to T3 (ANCOVA, unadjusted).

*p< 0.05 between groups at baseline.

NT pro-BNP

The number of available blood samples were at baseline (T1); 78, at T2 70 and at T3 69. One of the missing samples was due to one subject who presented with NT pro-BNP concentrations below the detection limits for the assays (<10 pmol/l). As shown in Table II, the NT pro-BNP levels did not differ between the groups at baseline. There were no significant differences in change between groups from baseline to T2 (p=0.22) or from baseline to T3 (p=0.38). A spaghetti plot for each NT pro-BNP value is presented in the control group and the exercise group (Figure 1), at each time point (T1, T2 and T3), connected with straight line segments.

Figure 1. Individual NT pro-BNP levels from the patients in the control group and in the exercise group at baseline (Time 1) after four months (Time 2) and 12 months (Time 3) each time point.

There were also no significant changes from baseline to either post exercise (T2) or long-term follow-up (T3) in the NT pro-BNP levels within the groups.

Functional capacity

The effects of the exercise program on functional capacity have previously been reported in details (n=80) (12). In brief, as shown in Table II, there were no significant differences in functional capacity at baseline between the two groups. Functional capacity increased in the exercise group from baseline to T2 as compared to the control group (6MWT; p<0.001, cycle exercise time; p<0.001). After 12 months, the differences in improvements were still statistically significant in favor of the exercise group.

Correlations

At baseline correlation analysis on corresponding variables of NT pro-BNP levels, 6MWT, cycle exercise time and ejection fraction were performed. There were significant inverse correlations between the NT pro-BNP levels and the 6MWT (r=−0.24, p=0.035) (Figure 2) and the cycle exercise time (r=−0.48, p<0.001) (Figure 3). An inverse correlation was also observed between NT pro-BNP and left ventricular ejection fraction (r=−0.27, p=0.018). There was, however, no significant correlation between the change in NT pro-BNP and changes in the 6MWT (r=0.12, p=0.33) or the cycle exercise time (r=0.04, p=0.72) after the intervention (T2) in the whole study group. After 12 months (T3) there was still no significant correlation between change in the NT pro-BNP and either change in the 6MWT (r=−0.16, p=0.9) or change in the cycle exercise time (r=−0.14, p=0.24).

Figure 2. Correlation between the NT pro-BNP levels and the six minute walk test (meter) at baseline in the investigated population (n=78). r= −0.24, p=0.035.

Figure 3. Correlation between the NT pro-BNP levels and the cycle exercise time (in seconds) at baseline in the investigated population (n=78). r=−0.48, p<0.001.

Discussion

Our study demonstrated that despite highly significant improvement in functional capacity of group-based high-intensity interval training in optimally medical treated patients with CHF, no changes in the NT pro-BNP levels were found, although NT pro-BNP is considered the strongest neurohormonal prognostic indicator in CHF. In this population, the value of NT pro-BNP measurement in the evaluation of functional capacity seems to be limited, which means that aerobic interval training is unable to substantially influence the neurohormonal burden associated with CHF in a cohort of stable patients on optimal medical treatment. The significant inverse correlation between the NT pro-BNP levels and the measurements of functional capacity at baseline, is nevertheless in line with what have been reported by other investigators (15).

Several studies have been conducted, but they differ largely in the study population and design, duration and mode of exercise, and also in the outcome variable as measure of neurohormonal burden, BNP and NT pro-BNP. Thus, results reported are not easily comparable. Our results are in accordance with two recent published observational study from Arad et al. (15) and Prescott et al. (16) who both showed significantly improved walking distance without significantly change in the NT pro-BNP levels. Their population had similar baseline levels as our cohort, but the patients in the study from Arad et al. (15) had a substantially shorter walking distance at baseline. Also two other studies using BNP as the outcome measure, have failed to demonstrate any significant effect of exercise training on resting BNP levels (10,17). One additional study reported no correlation between the BNP levels and the 6MWT in a stable CHF population (NYHA II–IV) (18).

In other studies significant reduction of NT pro-BNP and/or BNP levels have been shown after exercise training, different from controls (9,19). The study of Conraads et al. (9), was, however, a non-randomized trial and their patients had substantially higher NT pro-BNP levels in the exercise group compared to their control group and also compared to our study population. The exercise programs were also different from our program.

Standard CHF medication has been shown to reduce BNP and NT pro-BNP levels substantially, however, with large inter-individual variation (20,21). Great biological variability obviously exists in these variables (22). In the study by Arad et al. (15), exercise training seemed, however, to disrupt the correlation between NT pro-BNP levels and functional capacity at baseline. It has been suggested that NT pro-BNP is not sufficiently sensitive to detect changes in functional capacity during clinical follow-up, due to its spontaneous variability, and that NT pro-BNP changes are more likely to be detected by medical intervention, which acts more directly at the myocardium (23). Four months of exercise training might be insufficient to translate into significant changes in NT pro-BNP in medically treated clinically stable patients. Alternatively, improvements in functional capacity might have been too small to influence the NT pro-BNP levels. However, the improvement of exercise time of 14% observed in our study is comparable or better compared with results from other randomized studies with exercise training in populations with CHF patients in NYHA-class II–III (1,4,11). In addition, exercise training in patients with CHF has been demonstrated to primarily affect peripheral working muscles (24), which leads to improvements in functional capacity, but may not perforce an effect on the NT pro-BNP levels. After all, the levels of NT pro-BNP in the present study were relatively low, and reduced levels would possibly be difficult to obtain.

Assessment of functional capacity is associated with great prognostic impact in patients with CHF, where peak oxygen consumption represents the gold standard. However, the 6MWT has been shown to be closely related to peak oxygen consumption (25) and to represent an independent predictor of morbidity and mortality (26). Although the number of patients included in this study was relatively small, our study support the notion that NT pro-BNP is not a reliable marker for measuring changes in functional capacity in a population with stable CHF in NYHA class II–III with optimal medical treatment.

Conclusion

In the present population, no significant changes in NT pro-BNP levels were observed after four months of exercise training in a group-based high-intensity aerobic interval-training program, despite significant improvement in functional capacity and quality of life.

Acknowledgements

We acknowledge the help from physicians Svein Solheim, Haakon Kiil Grøgaard and Torstein Jensen, Department of Cardiology, and technicians from the exercise laboratory at Department of Cardiology, Oslo University Hospital, Ullevaal, regarding the cycle ergometer tests. We also are indebted to the Department of Clinical Chemistry at Oslo University Hospital, Ullevaal, regarding the NT pro-BNP analysis. The present study was supported by grants from the Eastern Norwegian Health Authority, Norway and The Norwegian Foundation for Health & Rehabilitation, Oslo, Norway. There is no conflict of interest to be declared.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.