Changes in fatigability following intense aerobic exercise training in patients with interstitial lung disease

Summary

Objective:

To determine if, in patients with interstitial lung disease (ILD), fatigue might be lessened after vigorous aerobic exercise.

Methods:

13 physically inactive patients (5 men and 8 women; age 57.2 ± 9.1 years, BMI 28.2 ± 4.6 kgm⁻²) with ILD of heterogeneous etiology and able to walk on a motor driven treadmill without physical limitation were enrolled. Subjects underwent cardiopulmonary exercise (CPET) and 6-min walk (6MWT) tests and completed Fatigue Severity Scale and Human Activity Profile questionnaires before and after an aerobic exercise-training regimen. The training regimen required participation in at least 24 of 30 prescribed aerobic exercise training sessions at a target heart rate of 70–80% of the heart rate reserve, 30 min per session, 3 times per week for 10 weeks.

Results:

After training, a 55% (p < 0.001) increase in time to anaerobic threshold on the CPET, and an 11% (p = 0.045) reduction in performance fatigability index (PFI), calculated from the performance on the 6MWT were observed. Distance walked on the 6MWT (6MWD) increased by 49.7 ± 46.9 m (p = 0.002). Significant improvements in scores on the Fatigue Severity Scale (p = 0.046) and Human Activity Profile (AAS p = 0.024; MAS p = 0.029) were also observed. No adverse events related to the training regimen were noted.

Conclusion:

After training, the decrease in fatigability appeared to result in increased 6MWD and was associated with physical activity. Since significant declines in 6MWD may be a marker for impending mortality in ILD, a better understanding of the etiological state of fatigue in patients with ILD and its reversal might provide fundamental insight into disease progression and even survival.

Introduction

Fatigue is a state of physiological decline in response to over exertion. The state manifests symptomatically as feelings of tiredness or weariness and can be observed as a decrease in physical or mental performance. Fatigability is a quantifiable phenotype in which the physiological decline is measured as a function of the duration, intensity or frequency of exertion, representing an important aspect of functionality. For example, hypothetical Groups A and B report similar levels of fatigue on a given instrument even though Group A reports participating in less physical activity. Group A would therefore appear to fatigue more easily and thus be more fatigable than Group B. Fatigability can be measured more definitively in the laboratory where the processes of physiological decline can be quantified as a response to controlled perturbations such as physical exercise. All humans are fatigable but fatigability is increased by poor physical fitness and becomes even more severe with medical conditions, such as interstitial lung disease (ILD) [1], in which cardiorespiratory function is impaired.

ILD is a condition resulting from over 200 etiologies [2–4]. Homogeneity within the condition results from diffuse and irreversible fibrotic repair of the alveolar parenchyma, regardless of the underlying etiology. A progressive decline in alveolar diffusion capacity and restriction of pulmonary ventilation ensues, which creates a gas exchange impediment that is often abetted by diminished alveolar blood flow resulting from capillary destruction and hypoxic vasoconstriction [2,5]. Exercise capacity is severely limited [1] and patients often find it difficult to sustain even moderate levels of physical functioning including instrumental activities of daily living and the ability to maintain employment [6].

The total distance that patients can walk (6MWD) on a 6-min walk test (6MWT) is a surrogate measure of exercise capacity [7–14]. Progressive decline in 6MWD may predict one-year mortality in patients with ILD, specifically those with idiopathic pulmonary fibrosis [14]. Clinically significant increases in 6MWD have been observed following aerobic exercise training in patients with ILD [7–13] and one study found the increase in 6MWD to be concurrent with a reduction in the severity of patient-reported fatigue [9].

In addition to significant increases in 6MWD, we have observed a novel pattern of adaptation to aerobic exercise training in patients with ILD [15]. The adaptation is characterized by an improvement in work economy despite the absence of increases in the peak capacity of the cardiorespiratory system. It therefore seems plausible that, fatigability might still be modulated by cardiorespiratory function, even though limitations on the maximum capacity of the system seem to be irreversible.

A more complete understanding of how fatigability is modulated could provide important insight into the progression of ILD, how fatigue severity might be diminished, and how treatment might be further advanced in general. Moreover, because fatigability is believed to be a factor contributing to 6MWD limitation, a greater understanding of its modulation may provide insight into improving clinical status and life expectancy in patients with ILD and perhaps those with advanced lung disease in the broader sense. In the current study we examined the hypothesis that variability in cardiorespiratory function could exert a modulating influence on fatigability that is independent of variability in the maximum capacity of the system.

Methods

Subjects

The authors’ respective institutional review boards approved this study and signed informed consent was attained from all participants prior to implementation of any of the study procedures or data collection. Subjects were recruited from local outpatient lung disease clinics within the Washington DC metropolitan area and enrolled in the National Institutes of Health Exercise Therapy for Advanced Lung Disease Trial [ClinicalTrials.gov identifier NCT00678821]. Subjects were required to be at least 21 years of age and physically inactive, defined as not having participated in any pulmonary rehabilitation program or structured exercise-training regimen of 30 min per session or longer, three or more days per week, during the 6 months prior to enrollment. None of the subjects had a history of ischemic heart disease, dilated, hypertrophic, or non-idiopathic cardiomyopathy, hepatic or renal dysfunction, severe psychiatric condition, use of antiretroviral therapy, uncontrolled diabetes mellitus, mitochondrial disease, acute cor pulmonale, disabling stroke, cancer, or evidence of these conditions at baseline physical examination. Pregnant women and potential subjects admitting to ongoing use of tobacco or abuse of alcohol or controlled or illegal substances were also excluded. Patients were eligible to enroll only if they had right ventricular systolic pressure (RVSP) ≤40 mmHg measured by echocardiogram within one year of enrollment, FEV₁/FVC ratio ≥65%, and ejection fraction ≥40%. Subjects were instructed to abstain from performing any exercise outside of the protocol procedures during their participation in the study.

Study design

After enrollment, subjects completed a cardiopulmonary exercise test (CPET), from which a target heart rate tor the exercise training sessions was determined, and a 6MWT from which the performance fatigability index (PFI) was calculated [16]. Subjects also completed the Fatigue Severity Scale and Human Activity Profile questionnaires to assess patient-reported fatigue severity and physical activity levels. Subjects then underwent a 10-week aerobic exercise program after which the CPET, 6MWT, Fatigue Severity Scale, and Human Activity Profile were repeated. Data acquired after training were compared to baseline data to determine post-training differences in fatigability, 6MWD, patient-reported fatigue severity, and physical activity.

Measurements

Cardiorespiratory exercise test

A detailed explanation of this procedure has been reported previously [15, 17]. Briefly, cardiorespiratory function was measured during graded treadmill exercise tests with continuous, breath-by-breath, pulmonary gas exchange measurements (CPET), ending in volitional exhaustion. A modified Naughton protocol was used. For this study our cardiorespiratory variable of interest was the time taken to attain the anaerobic threshold (AT). The oxygen consumption (VO₂) at which the AT occurred was determined by the V-slope method [18] and the time to AT was recorded.

Performance fatigability test

Performance fatigability is the decline in performance measured as a function of the duration, intensity, and/or frequency of exertion. PFI was derived from changes in velocity during the 6MWT, which was conducted in accordance with the American Thoracic Society (ATS) Guidelines on an 80-m, circular course [19]. Subjects were instructed to walk as far as possible during the test. Subjects requiring supplemental oxygen during exercise were provided with 6 l of oxygen per minute through a nasal cannula. PFI was calculated by first determining the average velocity over the entire six minutes of the 6MWT and the velocity for the first lap. The fractional change in walking velocity was then determined by dividing the 6-min velocity by the first lap velocity attained during the test. The PFI was finally calculated by normalizing the ratio of the fractional change in velocity to the total 6MWD.

Fatigue severity

The Fatigue Severity Scale [20] is a nine-item, self-scored questionnaire which uses a visual ranking format ranging from 1 to 7 to measure patient-perceived fatigue. The scores reported on the nine items are then averaged to create a composite score. Higher composite scores indicate more severe fatigue.

Physical activity

The Human Activity Profile is a self-report questionnaire that has been concurrently validated with respect to cardiopulmonary exercise [21]. Represented in the Profile are 94 questions on metabolic demands for a variety of life situations, ranging from social and physical functioning to activities of daily living such as self-care and occupational- related tasks [22]. Participants respond to each question as (1) still doing this activity, (2) have stopped doing this activity, or (3) never did this activity for each of the activities queried. The Human Activity Profile provides information in two subscales, the Maximum Activity Score (MAS) and the Adjusted Activity Score (AAS). The MAS represents the activity still being performed that requires the highest oxygen consumption, and is therefore the test’s indicator of an individual’s maximum capacity for performing physical activity. The AAS measures an individual’s current participation in all daily physical activities and tasks. Higher MAS and/or AAS scores are indicative of higher levels of physical activity.

Aerobic exercise training

Following completion of baseline testing, subjects participated in a vigorous aerobic training program 3 days per week for 10 weeks. Details of the training regimen were previously reported [15]. Briefly, subjects walked on a treadmill a velocity and inclination that elicited a heart rate response of between 70 and 80% of their heart rate reserve, as determined by the method of Karvonen [23], where peak heart rate was measured during the baseline CPET. Training heart rate was sustained for 30–45 min each session. Subjects requiring supplemental oxygen were provided the necessary flow rate to maintain SpO₂ ≥90%. Subjects were required to participate in at least 80% of the exercise sessions (24 out of 30 total sessions).

Statistics

The primary measures of interest were time to AT and PFI. Other outcome measures were 6MWD, and Fatigue Severity Scale score and The Human Activity Profile subscales, MAS and AAS. Dependent t-tests were used to determine if pre- and post-training scores differed. The significance of relationships among the dependent variables was examined by Pearson product moment correlation coefficients (r) and described by linear regression. Differences and correlations between the variables were significant when p-values were ≤0.05. Central tendencies are reported as the mean ± 1 standard deviation unit.

Results

Subjects of this study were 13 patients (Table 1) with ILD from a variety of etiologies and NYHA/WHO Functional Classes II and III as previously described [15]. The patients’ PFT and DLCO were consistent with a restrictive breathing pattern and impaired alveolar diffusion. Five of the 13 subjects (38.5%) received supplemental O₂ during the CPETs, 6MWTs, and exercise training sessions. All five of these subjects were on daily O₂ therapy. None of the subjects were awaiting lung transplantation. In total, subjects participated in 90% of the 30 exercise training sessions. No study-related adverse events occurred during either the exercise testing or training sessions.

Table 1.

Subject characteristics.

Subject	Age (yrs)	Sex (M/F)	BMI (kg/m2)	Etiology	FEV,/FVC (%)	DLCO (%pred)	mPAP (mmHg)	RVSP (mmHg)	EF (%)
1	54	F	32	NSIP	78	36	21	-	-
2	55	F	30	NSIP	91	35	-	30	60
3	71	M	32	IPF	85	45	21	-	-
4	56	M	21	NSIP	86	26	24	44.3	51
5	69	M	26	IPF	86	32	10	12	55
6	49	F	23	NSIP	83	-	16	37	60
7	42	M	25	DIP	80	33	-	35	65
8	69	F	28	Sj	84	29	18	34	60
9	66	F	25	IPF	84	79	-	27.8	63
10	52	F	28	SS	81	36	17.3	28	65
11	55	F	38	NSIP	86	33	-	-	55
12	47	F	29	SS	81	31	-	18	65
13	57	M	29	IPF	73	42	-	13	55
Mean	57.1	-	28.3	-	82.4	39.6	18.2	29.7	60.1
SD	9.1	-	4.4	-	4.7	14.7	4.5	11.6	4.8

FEV = Forced expiratory volume, FVC = Forced vital capacity, DLCO = Diffusion capacity of lungs for carbon monoxide, mPAP = Mean pulmonary artery pressure, RVSP = Right ventricular systolic pressure, EF = Ejection fraction, NSIP = Nonspecific interstitial pneumonitis, IPF = Idiopathic pulmonary fibrosis, DIP = Desquamative interstitial pneumonia, Sj = Sjogren’s syndrome, SS = Scleroderma/systemic sclerosis, SD = Standard deviation.

A more detailed description of cardiorespiratory function at baseline and after aerobic exercise training in these subjects was previously published [15]. Subjects’ cardiorespiratory fitness, as determined by peak VO2, was below the 10th percentile for people of their age and gender [24]. VO₂ at AT was not significantly above the minimal level required for sustaining the lowest intensities of instrumental activities of daily living (IADL) [25–27] and peakVO₂ was not significantly higher than the upper limit for IADL requirements, either before or after training (Fig. 1A). A significant difference in neither VO₂ at AT nor peak VO₂ was observed following training.

Figure 1.

Oxygen consumption (VO₂) at the anaerobic threshold and at peak exercise during CPET before (white) and after (black) training (Panel A). Gray bar in background represents the range of VO₂ requirements for the instrumental activities of daily living. Cardiorespiratory response pattern to CPET, depicting the change in time to anaerobic threshold (AT-time), observed in a representative subject (Panel B). For the group represented by this subject AT-time increased from 307.2 ± 122.9 s at baseline to 450.5 ± 158.2 s following training (p < 0.001).

Time to AT increased by 55% (p = 0.001) following training (Table 2). Measures obtained from the 6MWT were also improved after aerobic exercise training. PFI was 11% lower and 6MWD was almost 50 m longer in the trained condition than at baseline. 6MWD correlated significantly with both PFI (Fig. 2A) and time to AT (Fig. 2B). However, PFI, correlated with time to AT only after training (Fig. 2C).

Table 2.

CPET and performance fatigability before and after exercise training.

	Before	After	Difference	p-value
RHR (b/min)	101.8 ± 14.2	93.7 ± 12.8	−8.1	0.018
Peak HR (b/min)	147.8 ± 17.8	143.5 ± 13.9	−4.3	0.172
Peak VO2 (ml/kg/min)	17.6 ± 5.6	18.5 ± 5.0	+0.9	0.057
Peak WR (watts)	129.5 ± 56.2	170.5 ± 79.0	+41.0	0.002
CPET-time (sec)	736.8 ± 159.9	905.9 ± 256.3	+169.1	<0.001
AT-VO2 (ml/kg/min)	11.5 ± 3.4	12.8 ± 2.8	+1.3	0.095
AT-time (sec)	307.2 ± 122.9	450.5 ± 158.2	+143.3	<0.001
SpO2 nadir (%)	88.6 ± 7.4	87.4 ± 8.2	−1.2	0.069
Peak dyspnea-CPET (1 –10)	5.2 ± 2.2	4.3 ± 1.8	−0.9	0.075
Peak dyspnea-6MWT (1 –10)	3.3 ± 1.8	2.9 ± 1.3	−0.4	0.447
6MWD (meters)	433.0 ± 92.6	482.7 ± 92.8	+49.7	0.002
PFI	0.225 ± 0.064	0.199 ± 0.049	−0.026	0.045
FSS	4.1 ± 1.2	3.5 ± 0.9	−0.6	0.046
HAP-AAS	51.5 ± 12.4	59.6 ± 14.7	+8.1	0.024
HAP-MAS	64.1 ± 8.7	71.3 ± 10.0	+7.2	0.029

CPET = Cardiopulmonary exercise test, RHR = Resting heart rate, HR = Heart rate, VO₂ = Oxygen consumption, WR = Work rate, AT-time = Time to anaerobic threshold, AT-VO = VO₂ at anaerobic threshold, SpO₂ = Saturation of peripheral oxygen, 6MWT = Six-minute walk test, 6MWD = Six-minute walk test distance, PFI = Performance fatigability index, FSS = Fatigue severity scale, HAP = Human activity profile, AAS = Adjusted activity score, MAS = Maximum activity score.

Figure 2.

Correlations among the indices of performance and fatigability before and after aerobic exercise training. In Panel A, 6-min walk distance (6MWD) correlated significantly with the performance fatigability index (PFI) and in Panel B, with time to anaerobic threshold (AT-time), before and after training. In Panel C, PFI correlated with 6MWD only after training.

Subjects reported improvements on the Fatigue Severity Scale, as well as both subscales of the Human Activity Profile activity following training (Table 2). The Fatigue Severity Scale was unrelated to either PFI or time to AT (Fig. 3A). MAS was not associated with PFI and was related to time to AT only after training (Fig. 3B). Moreover, significant correlations were observed between AAS and both PFI and time to AT only after aerobic exercise training (Fig. 3C).

Figure 3.

Pre and post-training relationships among measures of fatigability and patient-reported fatigue severity and physical activity. In Panel A, Fatigue Severity Scale (FSS) was unrelated to either the performance fatigability index (PFI) or time to anaerobic threshold (AT-time). In Panel B, the Human activity Profile (HAP) Maximum Activity Score (MAS) was not related to PFI and was related to AT-time but only after training. In Panel C, HAP Adjusted Activity Score (AAS) was related to both PFI and AT-time but only after training.

Discussion

Fatigability was reduced following aerobic exercise training in these patients with ILD. The increase was evidenced by improvements in physiological and performance measures. Concurrent improvements in patient- perceived fatigue severity and patient-reported physical activity were also observed. Performance of vigorous physical exercise may be adversely influenced by the perception of tiredness or weariness (perceived fatigability). In the current study, perceived fatigability was measured neither during the 6MWT nor during CPET so its influence on performance could not be assessed. Conversely, physiological measures such as VO2 at AT and time to AT are not affected by motivation or the awareness of the symptoms of fatigue.

When physical activity becomes strenuous, increased production of high-energy phosphates is required for quenching the energy demand. ATP hydrolysis and anaerobic glycolysis are increased to levels that result in sarcoplasmic ion concentrations that exceed intracellular buffering capacity. A state of fatigue ensues due to ion competition with substrates for binding sites in the metabolic pathways, ion pumps and cross-bridge binding sites, restricting continuance of activity [28]. Metabolic acidemia is precipitated in an attempt to further buffer the ionic load, which results in non-metabolic CO₂ expiration during exhalation.

During CPET, CO₂ expiration rises linearly as a function of increases in VO₂. However, when the exercise becomes strenuous, the rise in CO₂ deflects upward more rapidly. The deflection point defines the gas exchange AT [18], making it a physiological marker of the onset of exercise-induced fatigue, albeit the onset of academia precedes the deflection point just slightly. VO₂ at AT is typically observed at approximately 40–60% of peak VO₂ in healthy adults [29–32], and is generally reduced in those whose cardiorespiratory fitness is poor [32]. VO₂ at AT is further diminished by medical conditions in which cardiorespiratory function is impaired [17,33–35]. VO₂ at AT can be increased by aerobic exercise training in most subsets of the general population [30,36–38] and is often very high in those with extremely high levels of cardiorespiratory fitness [39].

Perceptions of tiredness and reductions in performance are results of the state of fatigue. However, neither the physiological characteristics of the state nor its phenotype, fatigability, are influenced by them. Prolonged time to AT following aerobic exercise training has generally been attributed to increases in VO₂ at AT directly [30,36–38]. In the current study, time to AT was prolonged, despite the absence of training-induced increases in VO₂ at AT. This observation implicated processes, other than those that would directly effect increases in VO₂ at AT, as contributing to the reversal of fatigability in general [15,17].

In our previous report [15], the cardiorespiratory adaptation to vigorous aerobic exercise training was characterized by a rightward shift in the slope of work rate plotted on VO₂ (WR-VO₂ slope) during CPET. This finding was surprising in that it suggested an improvement in work economy had occurred without observable changes in maximum cardiorespiratory capacity as would normally be expected. The improved economy mediated longer time to AT and CPET duration and appeared to be mechanized by improved muscle O₂ extraction that occurred without changes in muscle O₂ availability and without improvement in VO₂ at AT. The available information therefore underscores muscle O₂ extraction capacity as a potential modulating factor for the severity of fatigability and suggests that, in patients with ILD, muscle oxygenation and metabolic economy could play significant roles in determining time to AT.

The method of measuring performance fatigability in the current study was a modification of the method of Schnelle and coworkers [16]. The method quantifies performance fatigability as a function of self-selected changes in exercise intensity during a timed walk test. The assessment was originally performed using a 10-min walk test (10MWT) in a group of elderly subjects, age 72–96 years (mean age 85 ± 6.9 years) residing in independent living facilities, who were clinically frail with slow gait speed. Test-retest stability of PFI was demonstrated by a very high correlation coefficient, although intraclass coefficients were not provided. PFI was associated with both patient-perceived fatigue severity and patient-reported physical activity.

Measurement of PFI in the current study was based on 6MWT performance. Tests were conducted in a circular rather than a linear corridor. The 6MWT is usually well tolerated by patients with ILD [7–14], as well as lung diseases in general [40–46]. Patients with ILD may find it difficult or impossible to sustain intense walking over longer distances or for prolonged durations. The 6MWT is the most frequently used measurement of exercise capacity in patients with ILD [7–14] and is easily interpreted by nearly all clinicians who treat restrictive lung diseases. Results of the current study were similar to those of Schnelle and associates [16] in that PFI was strongly associated with timed walk test distance, as well as physical activity. In addition, our physiological measure of fatigability (time to AT) was indirectly associated with performance fatigability (PFI) after training, whereas neither of these fatigability measures was associated with patient-perceived fatigue severity. These findings again pointed toward an underlying physiological modulator for the post-training changes in fatigability and exercise performance.

Schnelle and coworkers employed a scale for measuring perceived fatigability over the course of the timed walk test used to measure exercise performance in that study [16]. Perceived fatigability was associated with PFI (r = 0.94) and gait speed (r = 0.47), providing evidence that both the measure of performance fatigability and walking velocity could have been influenced by symptoms of increasing tiredness. Perceived fatigability was not measured during the 6MWT in the current study. Thus, it remains difficult to determine the extent to which the symptoms of increasing fatigue influenced the way the subjects performed on the 6MWT either before or after training.

As with many studies of advanced lung disease, our sample was small and heterogeneous with respect to etiology. While the sample cannot be considered as representative of the ILD population at large, its size may reflect the low prevalence of ILD in the general population [47]. Nevertheless, significant reductions in fatigability appeared to result in increased 6MWD and patient-reported physical activity in this subset. Moreover, cardiorespiratory function and fatigability were linked mechanistically.

The clinical course of ILD may be characterized by progressive or stepwise declines in 6MWD and exercise capacity. In the current study, a non-exercising or sham-exercising control group was not included so declines in 6MWD could not be detected over the 10-week course of the training regimen. While spontaneous improvement was unlikely, the training-induced decreases in fatigability and the increase in 6MWD might possibly have offset undetected declines that could otherwise have occurred had these subjects not participated in the exercise-training regimen.

Clinical implications

Fatigue is debilitating in patients with ILD and diminishes participation in physical activity [1,6]. In the current study, the decrease in fatigability, occurring after a vigorous aerobic training regimen, resulted in substantial increases in exercise capacity and physical activity. Although the exercise training intensity was higher than in most previous studies [7–13], the sessions were well tolerated with a high degree of attendance and participation in the exercise sessions.

Overall, the results of this study provide evidence that aerobic exercise training might be useful as an effective intervention on both the mechanisms and consequences of fatigability in patients with ILD. The training regimen was effective in the current study for patients with ILD of heterogenic etiology, who present with impaired lung diffusion capacity and normal ejection fraction. In a previous study, we reported that a similar aerobic exercise training regimen resulted in improved cardiorespiratory function and 6MWD in patients who had pulmonary artery hypertension in addition to ILD [17]. Questions remain regarding optimal initiation of exercise training in patients with ILD and its effectiveness or safety for patients with poor baseline 6MWD of less than approximately 240 m. Aerobic exercise carries with it little financial burden and is generally safe, especially when medically recommended. Aerobic exercise can be made available and accessible to nearly all patients with ILD, even as a primary adjunct therapy, as soon as its safety has been determined.

Conclusion

Fatigability was reduced following aerobic exercise training in these patients with ILD. The reduction in fatigability was accompanied by improvements in patient-perceived fatigue severity and associated with patient-reported physical activity, and increased exercise capacity as measured by 6MWT. Cardiorespiratory function was physiologically linked, as a modulating influence, to fatigability. A greater understanding of the mechanisms underlying fatigability and the significance of its relative contribution to exercise capacity may be of paramount importance for further elucidating the role of aerobic exercise as a viable intervention in patients with ILD. Our observations provide a foundation and basis for further mechanistic and efficacy studies in larger ILD cohorts.