Reproducibility of Ventilatory Parameters, Dynamic Hyperinflation, and Performance in the Glittre-ADL Test in COPD Patients

ABSTRACT

Background: This study aimed to investigate the reproducibility of the ventilatory parameters and dynamic hyperinflation (DH) induced by the Glittre ADL-test (TGlittre) in chronic obstructive pulmonary disease (COPD) patients. Methods: Twenty-three patients with COPD underwent anthropometry, spirometry, TGlittre (TGlittre1 and TGlittre2) and pre- and post-test slow vital capacity. During the tests the ventilatory response was evaluated. Results: Regarding the TGlittre reproducibility, 87% of patients had a better performance in TGlittre2, and reduced on average 0.34 minute ± 0.62 (p = 0.01) from TGlittre1 time to TGlittre2, showing a learning effect of 6.34%. The difference average between tests was correlated with the time spent in TGlittre1 (r = −0.52; p < 0.05). The TGlittre time was statistically reproducible (intraclass correlation coefficient = 0.97; p < 0.001). Final ventilation parameters and their variations presented low-to-high reproducibility, except respiratory rate. The DH was similar (p > 0.05) in both tests, with low reproducibility in percentage, while in liters it was not reproducible. Conclusions: The TGlittre time and ventilatory parameters are reproducible, while DH is variable in COPD patients. A 6−7% learning effect was shown, and it is recommended to perform two tests.

KEYWORDS:

• Pulmonary disease
• chronic obstructive
• reproducibility of results
• activities of daily living
• outcome assessment

Introduction

Ventilatory limitation is often the main cause of limitation to exercise and activities of daily living (ADLs) in patients with chronic obstructive pulmonary disease (COPD) (1). It is characterized by the low ventilatory reserve and reduced inspiratory reserve volume (1, 2). During exercise and increase in ventilatory demand there may be increased air trapping, with consequent reduction in inspiratory capacity (IC) (1). This phenomenon is known as dynamic hyperinflation (DH) and is usual in COPD patients (1, 2). This results in an increased sensation of dyspnea, causing patients to perform fewer physical ADLs compared to healthy individuals (3).

In patients with COPD, the functional status deterioration has a significant association with morbidity and mortality (4), as well as frequency of exacerbations and hospital admissions (5), therefore the evaluation of this outcome is an essential part of pulmonary rehabilitation programs (6). In this context, the Glittre-ADL test (TGlittre) was developed in order to assess limiting ADLs reported by patients with COPD: sit and rise from a chair, walk up and down stairs, weight-bearing arm movements, and walking. It is a valid, reproducible, and reliable test that is responsive to pulmonary rehabilitation programs; however, information about the test's learning effects are still contradictory (7, 8).

Skumlien et al. (7) found a learning effect of 7%, while Tufanin et al. (8) found no statistically significant differences in performance time between two TGlittre with a 30-minute interval or between the physiological adjustments during the tests. Therefore, it is necessary to confirm whether the performance of only one TGlittre is sufficient to assess functional limitation in patients with COPD or if there is a learning effect that justifies conducting two tests. Furthermore, although Tufanin et al. (8) found no significant difference between the two TGlittre for variable minute ventilation (VE) and respiratory rate (RR), the reproducibility of additional ventilatory variables and DH during two tests has not been investigated. As DH is one of the main factors responsible for the dyspnea sensation during ADLs in patients with COPD (3), it is important to consider whether running a second test after a short rest period would induce a greater ventilatory demand and greater DH.

Thus, the aim of this study was to assess the reproducibility of the performance, the ventilatory response, and the DH of two TGlittre applied on the same day in patients with COPD.

Methods

Subjects

The study included patients with COPD (Global Initiative for Chronic Obstructive Lung Disease [GOLD 2–4]) (9) recruited from public hospitals and private clinics of the Greater Florianópolis area, Brazil. The inclusion criteria were as follows: COPD clinic diagnosis (confirmed by spirometry); smoking history of more than 20 pack-years; clinical stability in the last month prior to the beginning of the protocol; and age between 40 and 80 years. The exclusion criteria were the following: home oxygen therapy; current smoking or cessation for less than six months; associated diseases that could influence the study assessments (cardiomyopathy, musculoskeletal diseases, cancer, tuberculosis, and asthma); use of orthopedic prostheses and inability to perform any of the study evaluations. In order to ensure the absence of asthma, patients were asked about characteristic respiratory symptoms of the disease according to the criteria of Global Initiative for Asthma (GINA) (10). In addition, we excluded those with a history of asthma self-reported or reported by the doctor. Furthermore, patients who had exacerbation of the disease at the time of the protocol or who had completed a pulmonary rehabilitation program less than one year prior to inclusion in the study were also excluded. The study was approved by the Research Ethics Committee of the University of the State of Santa Catarina (CAAE: 07397212.3.0000.0118), and all participants signed the informed consent form.

Protocol

This was a cross-sectional study carried out in two days. On the first day, the participants underwent anthropometric measurements and a pulmonary function test. On the second day, they conducted two TGlittre tests (TGlittre1 and TGlittre2) with a 30-minute interval between them to allow vital signs, dyspnea sensation, and IC to return to baseline levels. All evaluations were performed in the morning.

Lung function

The pulmonary function test was performed with the EasyOne spirometer (NDD Medical Technologies, Zurich, Switzerland), previously calibrated on each day of assessment. American Thoracic Society/European Respiratory Society (ATS/ERS) methods and recommendations were followed regarding the reproducibility of acceptability criteria (11) and the reference values were based on the reference equation by Pereira, Sato, and Rodrigues (12).

Glittre-ADL test

The TGlittre consists in completing a circuit carrying a backpack (2.5 kg for women and 5.0 kg for men). The subjects started in the sitting position and then walked down a level hallway (total distance of 10 m). At the midpoint of the hallway, there was a staircase with two steps up and two steps down (17 cm high and 27 cm wide each). The subjects would then come to a shelf and had to move three objects weighing 1 kg each and positioned on the top shelf (at the height of their shoulders). They had to move the objects, one at a time, to the bottom shelf (waist-high) and then transfer them to the ground. Next, the objects had to be placed on the bottom shelf again and finally returned to the top shelf. The subject then had to go back the same way and sit on the chair again (starting position), repeating the ADL circuit immediately without any verbal encouragement throughout the test. The test consisted of five repetitions to be completed in the shortest time possible (7). Two tests were conducted at the same time of day and the same rater conducted both tests for the entire sample. The total time of each test was used for analysis.

Inspiratory capacity (IC)

Before and immediately after the end of the test, the slow vital capacity test (SVC) was performed according to ATS/ERS recommendations (11), with the patient seated and using the EasyOne portable spirometer (NDD Medical Technologies, Zurich, Switzerland). A minimum of three and maximum of 8 attempts were accepted and the reproducibility was based on two maneuvers with variation of less than 5% or 150 ml (13−16). The highest values of reproducible maneuvers were considered for analysis. DH was defined as a reduction in IC of at least 10% and/or 150 ml from baseline (15). The IC performance after the tests was timed to ensure that the measurements were taken within five minutes of the test completion (17).

Ventilatory parameters assessment

During the two TGlittre, the subjects used a portable spirometer (Spiropalm 6MWT; Cosmed, Rome, Italy), comprised of a data insertion/storage device, a flowmeter, a face mask, and an oximeter. This device captured the following variables: minute ventilation (VE), respiratory rate (RR), inspiratory time, expiratory time, and total respiratory time (T_insp, T_exp, T_tot), inspiratory volume (V_insp), ventilatory demand (VE/maximum voluntary ventilation [MVV]), heart rate, and oxygen saturation. The MVV was determined using the equation: FEV¹ × 37.5 (14). All values were recorded automatically by the system every 15 seconds during TGlittre. The equipment calibration was performed daily according to the manufacturer's recommendations. The same evaluator was responsible for collecting the ventilation parameters of the entire sample.

Sample size calculation

The sample size calculation was based on an expected intraclass correlation coefficient (ICC) of 0.70 for the TGlittre time, ventilatory variables, and DH, α = 0.05, and β = 0.10, yielding a size sample of 17 patients (18).

Statistical analysis

The study data, presented as mean and standard deviation, were submitted to the Shapiro–Wilk normality test, followed by the paired Student t-test or Wilcoxon test if nonparametric to compare baseline and final ventilatory parameters and their variation in the two tests. The Bland–Altman plot was used to test the tendency and agreement of time and DH between TGlittre1 and TGlittre2. Spearman's correlation coefficient was used to verify the correlation between the TGlittre1 time and the time difference between the two tests (TGlittre2 - TGlittre1). To check the reproducibility of the TGlittre time and the ventilatory parameters, ICC was used. The following classification was used for the ICC values: low reproducibility (ICC < 0.40), moderate reproducibility (ICC ≤ 0.75), and high reproducibility (ICC > 0.75) (19). The data were organized using SPSS/Statistical Package for Social Sciences version 20.0 with a significance level of 5% (p < 0.05).

Results

Twenty-three COPD patients (16 males) completed the study. Demographic and lung function data are shown in Table 1. In both tests, the patients were similar at baseline, i.e., no statistically significant differences were found in T_insp, T_exp, T_insp/T_tot, V_insp, VE, RR, VE/MVV, or baseline IC between TGlittre1 and TGlittre2.

Table 1. Characteristics of the subjects.

	Mean ± SD n = 23
Age, years	65 ± 6
Weight, kg	70.0 ± 13.1
Height, m	1.64 ± 0.06
BMI, kg/m^2	25.7 ± 4.5
Smoking history, pack-years	61.1 ± 31.4
GOLD 2, n	7
GOLD 3, n	7
GOLD 4, n	9
FEV1/FVC, L	0.47 ± 0.10
FEV1, L	1.08 ± 0.45
FEV1, %pred.	37.2 ± 15.4
FVC, L	2.23 ± 0.64
FVC, %pred.	60.5 ± 16.6
MVVpred. L/min	40.6 ± 17.1

kg: kilogram; m: meters; BMI: body mass index; GOLD: Global Initiative for Chronic Obstructive Lung Disease; FEV₁: forced expiratory volume in first second; FVC: forced vital capacity; L: liters; %pred.: %predicted; MVVpred: maximal voluntary ventilation predicted.

Reproducibility of the Glittre-ADL test

Eighty-seven percent of the subjects performed better on the second test. Times improved by 20.4 s ± 37.2 (p = 0.01) on average, showing a learning effect of 6.34%. The variation in time between the two TGlittre correlated with the TGlittre1 time (r = −0.52; p < 0.05). In addition, seven subjects who improved their test time by more than 30 seconds in the TGlittre2 had a TGlitter1 time of more than four minutes. The Bland–Altman plot confirms that the subjects took less time to completed the TGlittre2 compared to the TGlittre1 and demonstrates the tendency for increased reductions in the TGlittre2 time in the patients who took longer to complete the TGlittre1 (Figure 1). The TGlittre times showed high reproducibility (ICC = 0.97; p < 0.001).

Figure 1. Bland–Altman plot of the difference between two TGlittre plotted against the mean value of the first (1) and second (2) TGlittre for the entire group of patients. The central dotted line corresponds to the average difference between two TGlittre (−0.34 minute), whereas the lower and upper dotted lines correspond to lower (−1.58 minute) and upper (0.90 min) limits of agreement, respectively.

Reproducibility of ventilatory parameters and dynamic hyperinflation

The final ventilatory parameters showed moderate to high reproducibility, except for RR, which showed poor reproducibility (Table 2). Furthermore, RR was statistically lower in the TGlittre2 compared to the TGlittre1 (−1.30 ± 5.19 ipm; p < 0.05). The other final ventilatory parameters were similar in both tests. The variation of ventilatory variables (Δ) during the two tests was similar, with low to high reproducibility, again with the exception of RR, which was not reproducible (Table 3). Despite this, there was no significant difference between tests for RR (p > 0.05).

Table 2. Reproducibility of final ventilatory parameters and performance in the TGlittre.

	TGlittre1 Mean ± SD	TGlittre2 Mean ± SD	p	ICC (95% CI)
Tinsp, ms	775 ± 180	815 ± 197	0.09	0.83 (0.63–0.92)^*
Texp, ms	1235 ± 214	1246 ± 253	0.37	0.69 (0.39–0.86)^*
Tinsp/Ttot rate	0.38 ± 0.06	0.40 ± 0.07	0.09	0.87 (0.71–0.94)^*
Vinsp, mL	1151 ± 352	1190 ± 315	0.09	0.95 (0.89–0.98)^*
VE, L/min	31.0 ± 8.5	30.4 ± 8.9	0.48	0.88 (0.73–0.95)^*
RR, bpm	30.6 ± 5.20	29.3 ± 4.76	0.03	0.45 (0.07–0.72)^*
VE/MVV	0.83 ± 0.23	0.82 ± 0.26	0.76	0.89 (0.76–0.95)^*
IC, L	1.96 ± 0.54	1.95 ± 0.46	0.96	0.90 (0.78–0.96)^*
BORG	3 [0 – 9]^#	3 [0 – 7]^#	<0.01	0,92 (0,84–0,96)^*
Time, min	5.36 ± 2.65	5.01 ± 2.79	0.01	0.97 (0.90–0.99)^*

TGlittre: Glittre ADL-test; ICC: intraclass correlation coefficient; CI: confidence interval; T_insp: inspiratory time; ms: milliseconds; T_exp: expiratory time; T_insp: inspiratory volume; mL: milliliters; V_inspmax: maximal inspiratory volume; V_exp: expiratory volume; VE: minute ventilation; L/min: liters per minute; RR: respiratory rate; bpm: breaths per minute; VE/MVV: ventilatory demand; IC: inspiratory capacity; L: liters; min: minutes.

^* p < 0.05 for ICC;, ^#: median [minimum–maximum]

Table 3. Reproducibility of changes in ventilatory parameters and DH for the TGlittre.

	TGlittre1 Mean ± SD	TGlittre2 Mean ± SD	P	ICC (95% CI)
ΔTinsp, ms	−96.6 ± 210	−164 ± 310	0.57	0.57 (0.23–0.79)^*
ΔTexp, ms	−409 ± 437	−521 ± 464	0.31	0.38 (−0.04–0.67)^*
ΔVinsp, mL	424 ± 301	418 ± 265	0.89	0.66 (0.35–0.84)^*
ΔVE, L/min	15.9 ± 6.70	16.3 ± 6.69	0.64	0.84 (0.65–0.93)^*
ΔRR, bpm	5.71 ± 6.67	6.07 ± 4.92	0.81	0.28 (−0.15–0.62)
ΔVE/MVV	0.41 ± 0.14	0.42 ± 0.15	0.75	0.64 (0.31–0.83)^*
DH, L	−0.20 ± 0.22	−0.22 ± 0.16	0.59	0.31 (−0.13–0.64)
DH, %	−9.35 ± 10.4	−10.5 ± 8.4	0.66	0.37 (−0.06–0.68)^*

TGlittre: Glittre ADL-test; Δ: change (final – rest values); ICC: intraclass correlation coefficient; CI: confidence interval; T_insp: inspiratory time; ms: milliseconds; T_exp: expiratory time; T_insp: inspiratory volume; mL: milliliters; V_inspmax: maximal inspiratory volume; V_exp: expiratory volume; VE: minute ventilation; L/min: liters per minute; RR: respiratory rate; bpm: breaths per minute; MVV: maximal voluntary ventilation; DH: dynamic hyperinflation; L: liters.

^* p < 0.05 for ICC.

The DH induced by both tests was similar in both liters and percentage (p > 0.05); however, the DH was not reproducible in liters and the percentage DH showed low reproducibility (Table 3). The Bland–Altman plot shows high variability for DH between the two tests, with no tendency for decrease or increase in the second test compared to the first (Figure 2).

Figure 2. Bland–Altman plot of the difference between the DH of the two TGlittre plotted against the mean value of the first (1) and second (2) TGlittre for the entire group of patients. The central dotted line corresponds to the average difference between two DH (−0.18 L), whereas the lower and upper dotted lines correspond to lower (−1.26 L) and upper (0.90 L) limits of agreement, respectively.

Discussion

The main findings of this study demonstrate that the TGlittre time is reproducible, as are ventilatory parameters and their variations during the two tests, except for RR and DH. Some studies have used the TGlittre as a tool to evaluate the functional capacity of patients with pulmonary disease (7,20−24) and chronic heart disease (25). However, to date, only three studies evaluated the reproducibility and learning effect of TGlittre time in patients with COPD, with conflicting results. Skumlien et al. (7) found good reproducibility for the TGlittre in COPD patients and a learning effect of 7% after performing the two tests on different days. José et al. (23) performed two tests on the same day and also found good reproducibility (ICC: 0.91) for the TGlittre in hospitalized patients with exacerbation of chronic respiratory diseases, including COPD, and a learning effect of 17%. Tufanin et al. (8) found no statistically significant difference between the two TGlittre times, performed on the same day with a 30-minute interval. The results of this study showed a time reduction of approximately 6% in the second test compared to the first. In addition, patients who performed worse in the first test showed a greater learning effect. Hernandes et al. (26) observed the same behavior in the six-minute walk test in patients with COPD and found that a distance of less than 350 m in the first test was a determining factor for a clinically significant improvement (greater than 25 meters) in the second test. A possible explanation for this behavior is the fact that more functionally compromised patients do not perform the first test at maximum capacity for fear of dyspnea or fatigue. Thus, besides the learning effect after the first test, these patients may feel more secure and motivated to perform the second test. In addition to this, individuals with longer times in the first test subsequently have a wider range for variation (27, 28).

Tufanin et al. (8) suggest that, in cases where the application of two tests is difficult, it is possible to apply a single test because they found no differences in execution time between the two tests. However, it is important to emphasize that this study did not include very severe patients. In addition, the average time for the TGlittre1 was 4.35 minutes, less than that of the present study. Although the TGlittre still does not have a reference equation and clinically significant minimum difference, it is believed that a single test may underestimate the functional status of patients.

Of the studies that evaluated the reproducibility of the TGlittre, only the one by Tufanin et al. (8) verified the reproducibility of the ventilatory adjustments in patients with COPD. However, this measure was not tested with ICC and only the VE and RR were analyzed, with no differences between the two tests. In the present study, other ventilatory variables were analyzed because they are closely related to mechanisms that limit exercise in patients with COPD. It is known that these patients are unable to sufficiently increase ventilation to meet metabolic demands, even in short activities (29). In addition to the RR, inspired and expired volumes as well as expiratory time indicate the behavior of the final expiratory volume; therefore, they are relevant to understanding the development of DH. Because it is considered one of the main factors responsible for the worsening of respiratory effort and subsequent dyspnea in ADLs in patients with COPD (3), it is important to ensure that the execution of a second test within a short period of time does not lead to greater ventilatory demand and greater magnitude of DH. The results of this study demonstrate that the ventilation parameters behave similarly in both tests and that, although RR and DH are not reproducible, their values were not higher in the second test. Additionally, the final RR was significantly lower in the second test, while the DH showed great variability, with no tendency to increase or decrease, although the average bias for DH in tests 1 and 2 was −0.18 L, which is slightly above the threshold for DH (a reduction of 0.15 L). Furthermore, the 30-minute rest between tests was sufficient for patients to recover their baseline ventilatory values.

The reproducibility of ventilation parameters using ICC has been investigated previously in endurance tests on the cycle ergometer in a large sample of patients with COPD. Although the re-test endurance time was significantly higher than the first test, both test times and ventilatory variables (VE, RR, IC, tidal volume, and inspiratory reserve volume) were reproducible (30).

Unlike other tests that include a single motor task, the TGlittre involves multiple tasks that require the recruitment of several muscle groups at different intensities and for different amounts of time. Just in the activities in front of the shelf, patients spend about 50% of the total test time (21). The lack of reproducibility of RR and DH might be associated with the fact that, in the first test, patients identify the activities that are more difficult and develop ventilation strategies to be able to perform the second test with less dyspnea and fatigue. Furthermore, although the Borg scale has not been shown to be reproducible, the dyspnea shown by patients in the second TGlittre was statistically lower than the first, in which patients showed better performance.

The results of this study demonstrate that the administration of two TGlittre can be a more reliable assessment of the functional capacity of patients with COPD, especially those with greater functional impairment. Furthermore, the two tests can be performed on the same day with a 30-minute interval, without causing greater ventilatory demand and DH in the second test. It is also know that tests with multiple tasks as the TGlittre are more representative of ADL (31). A limitation of the present study may have been the failure to measure static lung volumes, which could have enhanced the assessment of pulmonary function and static hyperinflation. However, this does not compromise our results, since the main focus is the evaluation of dynamic lung volumes.

As far as we know, this is the first study to investigate the reproducibility of two TGlittre applied on the same day, focusing on the ventilatory response and using ICC as a statistical tool. The methodological strategy was carefully designed to ensure the reliability of the data: the same examiner administered both tests for the entire sample and was responsible for collecting the ventilation parameters; IC time after the tests was timed to ensure that the measures were carried out before the reduction in DH; and the tests were performed at the same time of day.

Conclusions

In summary, the TGlittre times and the ventilatory parameters in the test are reproducible, while DH is variable in patients with moderate to very severe COPD. The test has a learning effect of 6.34%, therefore two tests are recommended for a more accurate assessment of the functional capacity of these patients.

Declaration of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article, and also received no financial support for the research, authorship, and/or publication of this article.