© 2002 American Thoracic Society
Dyspnea and Decreased Variability of Breathing in Patients with Restrictive Lung DiseaseDivision of Pulmonary and Critical Care Medicine, Edward Hines Jr., Veterans Affairs Hospital; and Loyola University of Chicago, Stritch School of Medicine, Hines, Illinois Correspondence and requests for reprints should be addressed to Martin J. Tobin, M.D., Division of Pulmonary and Critical Care Medicine, Edward Hines Jr., Veterans Affairs Hospital, Route 111N, Hines, IL 60141. E-mail: mtobin2{at}lumc.edu
Patients with restrictive lung disease are typically dyspneic and have an increase in overall respiratory center drive, as a result of increased lung elasticity. When we subjected healthy volunteers to external elastic loads, their variability of breathing was lessened. Accordingly, we hypothesized that patients with restrictive lung disease display decreased variability of breathing and, also, that decreased variability of breathing is related to dyspnea. Breathing pattern was measured nonobtrusively over 1 hour in 10 patients with restrictive lung disease and in 7 healthy subjects. On a separate occasion, dyspnea was measured while all subjects copied different tidal volumes and frequencies. Compared with healthy subjects, the random fraction of breath variability was reduced in patients with restrictive lung disease: 27 times for expiratory time, 12 times for tidal volume, and 6 times for inspiratory time (p < 0.01 in each instance). Conversely, the nonrandom, correlated fraction for tidal volume was increased almost 3-fold in the patients (p < 0.01). Small variations from average resting tidal volume caused marked increases in dyspnea in patients, and the relationship was parabolic (r 2 = 0.97; p < 0.001). In conclusion, patients with restrictive lung disease adopt a tightly constrained breathing pattern, probably as a strategy for avoiding dyspnea.
Key Words: control of breathing • respiratory sensation • pulmonary fibrosis
Abnormalities in the control of breathing contribute to a number of clinical manifestations. A decrease in average respiratory center output causes hypercapnia, whereas an increase can cause dyspnea. The purpose of the respiratory controller, however, is not only to achieve a particular overall output, but also to carefully adjust output on a breath-to-breath basis (1, 2). Research on control of breathing has been based largely on measures of average output over some period of time, and little attention has been paid to mechanisms that determine constancy of output. Especially scant is information on how breath-to-breath variation in controller output might contribute to clinical manifestations of lung disease (1, 3–9). Simple measurements, such as coefficients of variation, indicate that healthy subjects display considerable variation in tidal volume and respiratory cycle time from one breath to the next (10). Signal analysis techniques reveal that this variability is composed of random and nonrandom fractions (11–16). The predominantly random character of the variability makes it possible for the respiratory system to engage in tasks other than gas exchange, such as speaking (11–16). A smaller fraction of variability is nonrandom, and the tidal volume and respiratory cycle time of a breath are significantly related to those of the preceding breath (11–14). When healthy subjects are faced with external chemical or mechanical loads, however, the random fraction decreases significantly (12–14). This decrease in random variability may lessen the freedom of the respiratory system to undertake behavioral tasks (3, 12, 14). Patients with restrictive lung disease are typically dyspneic and have an increase in overall respiratory center drive, which appears to result from increased lung elasticity (17–20). In healthy volunteers subjected to 2- and 3-fold increases in external elastic loads, we found a decrease in random variability of tidal volume and expiratory time at the higher load (12). Accordingly, we hypothesized that patients with restrictive lung disease display a decrease in the random breath-to-breath variation of breath components. To determine whether patients with restrictive lung disease select a resting breathing pattern that minimizes dyspnea (4, 21, 22), we separately measured dyspnea while the subjects copied various respiratory patterns. We hypothesized that the relationship between tidal volume and dyspnea would take a U-shape in the patients, that is, large increases in dyspnea would result from small variations (either increases or decreases) in tidal volume.
Subjects Ten men with restrictive lung disease, mean age of 58 years, and seven healthy age-matched male volunteers, mean age of 60 years, participated in this study (Table 1). Appropriate institutional review board approval and written informed consent were obtained.
Experimental Protocol Ventilation was measured nonobtrusively with an inductive plethysmograph during 1 hour of resting wakefulness (23). Additional details on ventilation measurements are available in the online data supplement. At the end of the hour, mean tidal volume and respiratory cycle time were calculated for each subject. These values were then displayed on a computer monitor and tracked by each subject. Each subject tracked nine different volumes, consisting of the mean baseline tidal volume and four volumes below and four volumes above the mean (50, 62.5, 75, 87.5, 125, 150, 175, and 200% of baseline). For each volume, respiratory frequency was adjusted to maintain constant minute ventilation. The volumes were imposed in stepwise increments or decrements, and the initial direction of the change was randomly assigned. The subject could see the tracing generated by his tracked breath and when it exceeded the target value, indicated by a horizontal line on the screen, he also received an auditory signal (beep). Subjects tracked each target value for 5 minutes, followed by 5 minutes of rest. At the end of each 5-minute tracking period, dyspnea was measured with a modified Borg scale using the question "how uncomfortable is your breathing?" To permit adaptation to the system, subjects practiced tracking for 10 minutes before the start of data acquisition.
Data Analysis
Accuracy of tracking
Accuracy was expressed as a percentage of the target value. The patients and control subjects tracked tidal volume with the same overall level of accuracy (p = 0.75) (see Figure E1 in the online data supplement). Both groups tracked each target of respiratory frequency with the same accuracy as at baseline, but the overall tracking of frequency was less accurate in the patients than in the control subjects (p < 0.05, analysis of variance [ANOVA]) (Figure E1).
Variability of resting breathing pattern
Mean Values and Gross Variability of Breath Components during Resting Breathing Mean values of inspiratory and expiratory times were shorter in the patients than in the control subjects (p < 0.05 in both instances); tidal volume and minute ventilation were similar in the two groups (Table 2).
The smaller breath-to-breath variation in tidal volume in a patient than in a healthy subject can be seen in Figure 1 . Gross variability of each breath component was strikingly reduced in the patients (Table 2). Compared with the control subjects, the patients displayed a 56% decrease in the coefficient of variation of tidal volume (p < 0.01), a 46% decrease in that of expiratory time (p < 0.05), and a 33% decrease in that of inspiratory time (p < 0.05) (Table 2).
Autocorrelation and Spectral Analysis Autocorrelation coefficients, which quantitate the relationship between values of a breath component in immediately adjoining breaths, were almost 2-fold higher for tidal volume (p < 0.01) and more than 2-fold higher for expiratory time (p < 0.05) in the patients as compared with the control subjects. The autocorrelation coefficients for the other components were similar in the two groups (Table 2). For each breath component, the power spectra were equivalent in the patients and in the control subjects.
Fractionation of Variational Activity of Breathing
Variation in Tidal Volume and Dyspnea At all levels of tracked tidal volume, dyspnea was greater in the patients than in the control subjects (p < 0.01) (Figure 3) . Dyspnea varied with tracked volume in the patients and the relationship fitted a parabolic regression: y = 10.3–0.12 x + 0.0005 x2 (r2 = 0.97, p < 0.001), where y = dyspnea score and x = tidal volume. Dyspnea did not vary with tidal volume in the healthy subjects.
Patients with restrictive lung disease showed a distinctive alteration in the variability of breathing, with a decrease in random variability and an increase in the correlated fraction (Figure 2 and Table 2). Slight variations of a patient's average tidal volume at rest provoked considerable dyspnea (Figure 3), and this vulnerability to dyspnea probably explains why the resting respiratory cycle is so tightly controlled in patients with restrictive lung disease. To our knowledge, this is the first report of systematic alterations in fractions of breath variability in a disease state and their relationship to clinical symptoms.
Variability of Breathing at Rest Speaking and eating, two of the major behavioral activities that involve the respiratory system, occur during expiration. Tidal volume and expiratory time have been shown to become more variable in healthy subjects during speech; inspiratory time showed a smaller increase in variability (25). Eating also causes an increase in the variability of tidal volume and expiratory time—but not in that of inspiratory time—in healthy subjects (26). The breath components primarily affected by speech and eating were the ones constrained the most in our patients: random variability in expiratory time and tidal volume were, respectively, only 4 and 8% of that in the healthy subjects, and the autocorrelation coefficients for expiratory time and tidal volume were, respectively, increased 2.3- and 1.9-fold. This curtailed ability to vary expiratory time and tidal volume—the components that become more variable during speech and eating—may contribute to the suffering experienced by patients with restrictive lung disease. Much insight into the pathophysiology of restrictive lung disease has been gained by studying healthy subjects breathing against external elastic loads (27). When we imposed 2- and 3-fold increase in elastic load in healthy subjects, we found that the higher load decreased the random variability in tidal volume and expiratory time (12). Of note, the decrease in variability with external loading was confined to the random fractions of tidal volume and expiratory time—the same fraction and breath components exhibiting the most striking abnormalities in our patients. The term "restrictive lung disease" was introduced to describe global decreases in the subsets of lung volume in these patients (28). The coinage was unconsciously prescient and fortuitous, because the term also correctly characterizes a second clinical manifestation: the restricted variation in the respiratory cycle from one breath to the next. The extent of the two abnormalities, however, differs by an order of magnitude. The diagnosis of restrictive lung disease requires a decrease in total lung capacity to 80% or less of the predicted normal value (29, 30)—it was 65% of that value in our patients. In contrast, random variability in expiratory time and tidal volume in our patients were, respectively, a mere 4 and 8% of the variability found in our healthy subjects.
Breathing Pattern and Dyspnea When lung elasticity is decreased, as in patients with restrictive lung disease, a deep breath will necessarily involve a large effort. Consequently, shallow breathing offers a strategy for avoiding dyspnea (31, 32). A deviation in tidal volume from the average resting value can cause dyspnea through stimulation of one of two sets of respiratory muscle receptors: low tidal volumes stimulate muscle spindles, and high tidal volumes stimulate tendon organs (33). A considerable body of data suggests that healthy subjects select the combination of tidal volume and frequency that minimizes respiratory effort (21, 22, 34, 35). The respiratory controller, however, is also responsible for ensuring the elimination of carbon dioxide (36, 37). Whereas a low tidal volume can decrease respiratory effort, the inevitable increase in dead space ventilation will cause hypercapnia and dyspnea (18, 37). A compromise between increased effort and clearance of carbon dioxide helps explain the parabolic relationship between variation in tidal volume and dyspnea in our patients and the decreased breath-to-breath variations in their resting breathing. Although the patients experienced marked increases in dyspnea with slight variations in tidal volume (Figure 3), they were able to track the imposed volumes with the same accuracy as the healthy subjects (Figure E1). That voluntary control of tidal volume was normal in the patients means that chemical and mechanical stimulation was not sufficient to interfere with their ability to copy the targets. In other words, the issue is not that the patients "can't" achieve a particular tidal volume during resting breathing because of mechanical or chemical constraints, but that they "won't" do it because of the resulting dyspnea (38). In summary, patients with restrictive lung disease display marked, and hitherto unrecognized, abnormalities in breath-to-breath variability of breathing; the random fraction of variability being as much as 27 times smaller than in healthy subjects, and the correlated fraction up to 3 times higher. Restricted variability was greatest for tidal volume and expiratory time—the components that vary the most during speech and eating. Slight variations from the average resting tidal volume caused large increases in dyspnea in the patients, but not in healthy subjects. In conclusion, patients with restrictive lung disease breathe in a very monotonous manner, probably as a deliberate strategy to avoid dyspnea.
The authors thank Malinda Mazur, Ranya Minawi, and Christopher R. Harley for technical assistance, Linda S. Fehr, M.S., for writing computer software, and Franco Laghi, M.D., for reviewing the manuscript.
Supported by grants from the Veterans Administration Merit Review, the Swiss National Science Foundation, and from the Swiss EMDO Foundation. This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form January 10, 2002; accepted in final form February 25, 2002
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
