|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Psychology (Drs. von Leupoldt and Dahme, Ms. Balewski, and Ms. Petersen), University of Hamburg, Hamburg; Atem-Reha GmbH (Dr. Taube and Mr. Schubert-Heukeshoven), Hamburg; and Pulmonary Research Institute at Hospital Grosshansdorf (Dr. Magnussen), Grosshansdorf, Germany.
Correspondence to: Andreas von Leupoldt, PhD, Department of Psychology, University of Hamburg, Von-Melle-Park 5, 20146 Hamburg, Germany; e-mail: andreas.vonleupoldt{at}uni-hamburg.de
Abstract
Background: Verbal descriptors of dyspnea are important in understanding the underlying mechanisms, but little is known about the language of dyspnea in COPD. We examined the language of dyspnea in COPD at different intensity levels of dyspnea.
Methods: Verbal descriptors of dyspnea were assessed in 64 patients with moderate-to-severe COPD (mean age, 62 years; mean percentage of predicted FEV1 [FEV1%pred], 54.1%) during slight dyspnea at rest (mean Borg score, 1.8), moderate dyspnea during cycle ergometer exercise (mean Borg score, 3.1) and somewhat severe dyspnea during a 6-min walking test before (mean Borg score, 4.2), and after pulmonary rehabilitation (PR) [mean Borg score, 3.5]. Furthermore, the influence of age, gender, baseline lung function (FEV1%pred), and PR on the verbal descriptors were studied.
Results: A cluster analysis showed that patients differentiated between five clusters of verbal descriptors of dyspnea: heavy/fast breathing, shallow breathing, obstruction, work/effort, and suffocation. These were related to the intensity level of dyspnea but not to age, gender, baseline lung function, or PR. While shallow breathing was predominant only during slight dyspnea at rest, heavy/fast breathing and to a lesser extent work/effort became more important during moderate and somewhat severe dyspnea during exercise. The clusters heavy/fast breathing and work/effort demonstrated the highest sensitivity in discriminating between different intensity levels of dyspnea and in characterizing the positive effects of PR.
Conclusions: Verbal descriptors of dyspnea in COPD are related to the intensity level of dyspnea. The clusters heavy/fast breathing and work/effort seem to be particularly sensitive descriptors of dyspnea during exercise in COPD.
Key Words: COPD dyspnea exercise perception qualitative descriptors
Dyspnea, the subjective experience of uncomfortable breathing, is an impairing symptom in many cardiorespiratory and neuromuscular diseases and the most prominent and limiting symptom in COPD.123 Dyspnea comprises of distinct sensations, which can vary in their quality and intensity.1 Based on the hypothesis that distinct qualities of perceived respiratory discomfort result from different pathophysiologic mechanisms, verbal descriptors of dyspnea have been suggested as being an important source of clinically relevant diagnostic information on the underlying pathophysiology, which might be of further importance for choosing the optimal treatment of this symptom.4567 Research891011121314 has demonstrated 2 up to 14 distinguishable qualities of dyspnea, which have partly been linked to specific underlying disease conditions. For example, patients with asthma preferred the terms chest tightness or work/effort to describe their feelings of dyspnea,5715 while patients with COPD used the terms work/effort, heavy or shallow breathing, and inspiratory difficulty.571617
However, the language of dyspnea in COPD is still insufficiently examined. A few studies5716 have investigated verbal descriptors of dyspnea in this patient group, mostly by using recall or by examining descriptors after one exercise condition. Thus, little is known about the language of dyspnea in COPD at different intensity levels of dyspnea. In asthma, however, it has been shown that the use of qualitative descriptors of dyspnea changes with increasing levels of respiratory loading.18
We therefore examined verbal descriptors of dyspnea in patients with COPD at different intensity levels of dyspnea during rest, cycle ergometer exercise, and 6-min walk test (6MWT). Since progressive changes in the respiratory and neuromuscular system in COPD during exercise are well known,2319 we hypothesized that these changes are reflected in the language patients use to describe their perceived dyspnea. Further, we examined the influence of age, gender, baseline lung function (percentage of predicted FEV1 [FEV1%pred]) and pulmonary rehabilitation (PR) on verbal descriptors of dyspnea because these factors have been shown to impact on the perceived intensity of dyspnea.2152021222324252627
Materials and Methods
Participants
We studied 64 patients with COPD (Table 1
) entering a 15-day outpatient PR program that included exercise training, patient education, nutrition counseling, breathing therapy, relaxation therapy, psychosocial education, and support for smoking cessation. All participants provided informed written consent. The study was performed at an outpatient PR center (Atem-Reha GmbH; Hamburg, Germany) and was approved by the local medical ethics committee.
|
Verbal Descriptors of Dyspnea
A German-language list of 22 respiratory symptom descriptors1314 adapted from Simon et al67 was presented after each experimental condition (Table 2
). Descriptors were rated on an 11-point Likert scale (0 = not at all, to 10 = very strong). Each participant was randomly assigned one of three different versions of the descriptor list; the versions differed only in the order of descriptors. In addition, the intensity level of dyspnea was rated after each condition on a separate modified Borg scale.30
|
At rest, participants were seated in a comfortable position on a chair for 2 min. This was followed by a 6MWT performed according to standards of the American Thoracic Society31 in a corridor 30 m in length. After a resting period of 60 min, participants performed incremental cycle ergometer exercise on an electronically braked cycle ergometer (Ergo1200cycle; ERGO-FIT; Pirmasens, Germany) at 5 W/min starting at 25W up to a symptom-limited maximum. Patients were instructed to stop exercising when the perceived symptom of dyspnea reached a Borg score of 1 U below the value of the 6MWT and were assisted by an experimenter. On the last day of the PR, patients performed a final 6MWT similar to that on PR entry day.
Data Analysis
Results are reported as mean ± SD. The descriptors of dyspnea were analyzed with a hierarchical cluster analysis. The aim of cluster analysis is to find homogeneous subgroups of items in the heterogeneous list of descriptors on the basis of their mutual similarities or distances with regard to various attributes.32 For this purpose, a distance matrix was computed indicating the distance of each item to each of the other items. All conditions (rest, cycle ergometer exercise, 6MWT) served as "attributes," and the distances were computed with reference to the ratings of all conditions simultaneously. The Ward algorithm with squared Euclidian distances was used, and results were confirmed by other fusion algorithms such as complete linkage, which indicates a stable cluster solution. Following this, the means of all items in each cluster were calculated for each condition. Cluster means, Borg scores, and distance walked during the 6MWT were analyzed with analyses of variance with a repeated-measures design and a Greenhouse-Geisser correction of degrees of freedom applied when appropriate. Based on median splits, age and FEV1%pred were included as between-subject variables in analysis of variance on cluster means (age
62 years vs age > 62 years, and FEV1%pred
50% vs FEV1%pred > 50%). Bonferroni-corrected, univariate, pairwise comparisons were calculated for further exploration of main effects. Gender differences in cluster means were analyzed for each condition with nonparametric Mann-Whitney rank sum tests. All analyses were calculated using statistical software (SPSS 13.0; SPSS; Chicago, IL) using a significance level of p < 0.05.
Results
Exercise Performance
Experimental conditions differed in their intensity level of dyspnea confirmed by a significant increase in Borg scores from rest (1.8 ± 1.5, slight dyspnea) to cycle ergometer exercise (3.1 ± 1.6, moderate dyspnea) to the 6MWT (4.2 ± 1.6, somewhat severe dyspnea) [all p < 0.001]. Compared to the first day, Borg scores at the 6MWT showed a significant decrease after PR (3.5 ± 1.3) paralleled by a significant increase in the respective distance covered in a 6MWT (457 ± 82 m and 493 ± 84 m) [p
0.001].
Verbal Descriptors of Dyspnea
Cluster analyses identified five clusters of verbal descriptors: (1) heavy/fast breathing, (2) shallow breathing, (3) obstruction, (4) work/effort, and (5) suffocation (Table 2), which showed different patterns depending on the intensity level of dyspnea (Fig 1
). During slight dyspnea at rest, the cluster mean for shallow breathing was greater than all other cluster means (p < 0.05). For moderate dyspnea during cycle exercise, the cluster mean for heavy/fast breathing was greater than all other cluster means (p < 0.001) except shallow breathing. Additionally, the cluster mean for shallow breathing was greater than that for the obstruction cluster (p < 0.01). For somewhat severe dyspnea during the 6MWT, the cluster mean for heavy/fast breathing was greater than all other cluster means (p
0.001) and the cluster mean for work/effort was greater than for obstruction (p < 0.01). The cluster mean for suffocation was smaller than all other cluster means in all conditions (p < 0.001). No differences in the cluster means were found between patients with FEV1%pred
50% and FEV1%pred > 50%, patients aged
62 years and aged > 62 years, or female and male patients in any of the conditions.
|
0.01) and from cycle exercise to 6MWT (p
0.01) [Fig 2
]. The cluster means for heavy/fast breathing and work/effort showed an additional increase from rest to moderate cycle exercise (p < 0.001 and p < 0.06, respectively) and the highest effect sizes (
2 = 0.43 and (
2 = 0.30, respectively) compared to cluster means for shallow breathing, obstruction and suffocation (
2 = 0.08, 0.15, and 0.11, respectively).
|
0.05). The cluster means for heavy/fast breathing and work/effort showed the strongest pre-PR/post-PR differences (
= 1.2 and 1.0, respectively) and the highest effect sizes (
2 = 0.29 and 0.21, respectively) compared to cluster means for shallow breathing, obstruction and suffocation (
= 0.6, 0.7, and 0.6;
2 = 0.08, 0.11 and 0.15, respectively). The pattern of cluster means for the 6MWT after PR was widely comparable to that of the first day, with the cluster mean for heavy/fast breathing being greater than all other cluster means (p
0.001), and that for suffocation being smaller than all other cluster means (p < 0.001). Discussion
We examined the language of dyspnea used by patients with moderate-to-severe COPD at a slight, moderate, and somewhat severe intensity level of dyspnea elicited by rest, cycle ergometer exercise, and 6MWTs. It was found that patients discriminated between five clusters of verbal descriptors of dyspnea: heavy/fast breathing, shallow breathing, obstruction, work/effort and suffocation, confirming previous findings that distinct qualities of dyspnea can be differentiated by patients with COPD5716 as well as by other patient groups579101215163334353637 and healthy individuals.68111314 The clusters of verbal descriptors used showed a distinct pattern depending on the intensity level of dyspnea. This was not influenced by age, gender, baseline lung function, or PR. The clusters heavy/fast breathing and work/effort demonstrated the highest sensitivity in discriminating between intensity levels of dyspnea and in characterizing the effects of PR.
The obtained clusters of German-language descriptors of dyspnea in patients with COPD in the present study converge with a number of previous studies56710111415 in various patient as well as healthy samples, mostly performed in English-speaking countries (eg, United States, United Kingdom, Canada). This indicates that these clusters are indeed distinct and separable cognitive constructs11 with a high degree of similarity, even across different languages and cultures. Moreover, the most prominent clusters observed in our study (heavy/fast breathing, work/effort, shallow breathing), conform with those observed in the few previous studies5716 in patients with COPD, while, importantly, not suffering the caveats of recall biases and undistinguishable intensity levels of dyspnea present in other studies, as previously outlined. During recall, patients in the study by Mahler et al5 exclusively selected descriptors of the cluster increased work/effort while patients investigated by Simon et al7 chose additionally the clusters heavy breathing, gasping, and hunger for air. Most of the descriptors in the latter two clusters were included in the clusters heavy/fast breathing and work/effort of the present study. Further, ODonnell et al16 demonstrated that patients with COPD selected the descriptors shallow breathing, work/effort, and heaviness of breathing to describe their breathing discomfort during incremental cycle ergometer exercise. However, descriptors denoting increased inspiratory difficulty and unsatisfied inspiratory effort that were additionally chosen by 75% of their patients played only a minor role in the current study, which might be related to the lower exercise level in our study.
The clusters heavy/fast breathing and work/effort demonstrated the highest sensitivity in discriminating between the three intensity levels of dyspnea by showing significant differences between all three conditions, the highest effect sizes for changes and the strongest differences in the pre-/post-PR comparison. This underlines the relevance of these specific language descriptors in patients with COPD, particular at the levels of exercise examined, at which states of hypoxemia and therefore descriptors such as suffocation are less common. The missing effects of age, gender, baseline lung function, and PR on the cluster pattern support this view and suggest that these descriptor clusters are stable cognitive constructs,11 which might be useful in assessing the level of deconditioning in patients with COPD.17 Particularly the strong decreases obtained in the clusters heavy/fast breathing and work/effort in combination with increased exercise performance in the 6MWT after PR support this idea.
Most importantly, we showed that verbal descriptors of dyspnea in COPD are related to the intensity level of dyspnea. Shallow breathing was the predominant descriptor cluster used only for slight dyspnea at rest and might have reflected mild ongoing hyperinflation as a consequence of, or in combination with, chronic expiratory flow limitation caused by destructive intrapulmonary changes.2319 These sensations are assumed to be sensed by mechanoreceptors in the chest wall or in the case of bronchoconstriction by intrapulmonary receptors mediated through vagal and autonomic pathways.1203839 For moderate and somewhat severe dyspnea during exercise, heavy/fast breathing and to a lesser degree work/effort became the most important descriptors. These might have mirrored the known consequences of increased ventilatory demand during increased levels of exercise in COPD, leading to acute dynamic hyperinflation and associated progressive mechanical burden and functional weakening of the respiratory muscles.23419 Thus, the changes in verbal descriptors of dyspnea obtained in patients with COPD during increased intensity of dyspnea might have reflected the involvement of different pathways mediating the underlying physiologic changes.42039
Similar findings have been reported by Moy et al18 for patients with asthma. In their study, patients chose chest tightness/constriction to characterize their breathing discomfort during mild methacholine-induced airway obstruction. At more intense levels of bronchoconstriction, the sense of work/effort progressively increased while chest tightness continuously decreased, which the authors18 interpreted with the recruitment of different underlying mechanisms and associated sensory pathways. Two further studies3435 in patients with asthma, however, found no difference in the verbal descriptors of dyspnea used at mild and at more intense levels of methacholine-induced airway obstruction, which might partly be related to the different degrees of hyperinflation induced in these studies.
There are some differences between the results of the current and aforementioned studies with regard to single descriptors included in the clusters. The descriptors "I feel out of breath" and "I am gasping for breath" for example are located in the clusters heavy/fast breathing and work/effort in the present study, while they form a single cluster (gasping) in the study by Simon et al.6 Further, the present study obtained five clusters of descriptors, whereas other studies reported between 2 distinguishable clusters89 and 14 fourteen distinguishable clusters7 or types of dyspnea. These differences are presumably due to methodologic differences. While some studies asked participants to recall qualities of dyspnea during exercise10 or activities,5 others questioned participants at rest about dissimilarities of pairs of descriptors11 or about the frequency of occurrence of specific dyspnea sensations over the past month.37 Yet other investigators6131436 presented their participants with different tasks to induce dyspnea, including resistive load breathing, exercise, or breathholding. Most obviously, different populations have been examined, ranging from healthy, young students61314 to severely disabled, older patients.5716 In combination with additional influences stemming from different languages, ethnic groups and cultures on participants perception of dyspnea,4373840 the above-named differences might have contributed to the discrepant findings of the reviewed studies.
A limitation of the present study is the lack of physiologic data during experimental conditions. These could have provided further insight into the relationships between specific descriptors of dyspnea, intensity levels of dyspnea, and underlying pathophysiologic mechanisms. In addition, studies4142 have shown that patients reported more dyspnea during walking than during cycling exercise. Therefore, we cannot exclude the possibility that cycle ergometer exercise not only produced a different intensity level of dyspnea, but could have also involved different pathways or pathomechanisms of dyspnea compared to the 6MWT and, thus, distinct qualities of dyspnea.
In summary, in the present study patients with COPD distinguished between five clusters of verbal descriptors of dyspnea that were related to the intensity level of dyspnea but not to age, gender, baseline lung function, or PR. The clusters heavy/fast breathing and work/effort were the most sensitive and important descriptors of dyspnea during exercise. Our results underline the importance of using specific verbal descriptors of dyspnea that can assist health-care professionals in understanding the language and mechanisms of dyspnea, and thus might help in optimizing the treatment of dyspnea.
Footnotes
Abbreviations: FEV1%pred = percentage of predicted FEV1; 6MWT = 6-min walk test; PR = pulmonary rehabilitation
The authors have no conflicts of interest to disclose.
Received for publication January 11, 2007. Accepted for publication March 26, 2007.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |