INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dyspnea is a common and unpleasant symptom in several respiratory, cardiovascular, and neuromuscular diseases and in normal subjects during exercise and neuropsychiatric hyperventilation. It has been attributed to a broad spectrum of stimuli grouped as mechanical, chemoreflex, or neurological.

The role of carbon dioxide (CO₂) on the generation of dyspnea has been studied since Haldane published his findings in the 1930s (1) and is still debated. When the chemoreceptors are activated as a consequence of hypercapnia, motor output and respiratory muscle activity increase, hyperventilation ensues, and muscle, joint, and pulmonary receptor activity increases. Each receptor type has the potential to generate conscious sensations and thus to play a role in the genesis of dyspnea.

It is evident that the dyspnea related to CO₂ could be a consequence of stimulation of the so-called respiratory centers and corollary copy to supramedullary structures and peripheral effects related to hyperventilation. Also, CO₂ could have a direct effect upon the central nervous system outside the chemoreceptors. These facts make it difficult to isolate the sensation resulting from chemoreceptor activation alone from that generated by muscular and/or pulmonary receptors.

Several studies have been dedicated to elucidating the role of CO₂ on dyspnea or related respiratory complaints. Earlier studies conducted during breath-holding (2, 3) and exercise (4) protocols showed relatively little role of CO₂ upon the associated distressing sensations.

Other studies (5), however, showed independent CO₂ effects upon respiratory sensations associated with breath-holding (11), exercise (8, 9), voluntary hyperventilation (9), CO₂ administration in chronically ventilated patients (5, 6) or conscious, paralyzed volunteers (7, 11) and intermittent CO₂ administration in patients with chronic obstructive pulmonary disease (COPD) (10).

In order to elucidate whether the inspiratory effort sensation (IES) associated with CO₂ is independent of the concomitant increase in the ventilation, we conducted the present study in normal, resting volunteers under CO₂ rebreathing. Our main goal was to compare the IES at a same hyperventilation level under hypercapnic and isocapnic conditions. Since the IES decreased under isocapnic hyperventilation, an independent CO₂ effect is suggested.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-three healthy subjects, (17 men and six women) selected from the personnel of the institute, mean age 34 ± 11 yr SD (range 24 to 73 yr) participated in the study. All were unaware of the scientific purposes of this study. All subjects were in good health without any history of pulmonary or neuromuscular disorders. Informed consent was obtained from all subjects before testing. The protocol of this study was approved by the Institutional Review Board for Human Study.

Breathing Circuit

Each subject was seated in a straight-back chair in a comfortable position. The subjects wore nose clips and breathed through a mouthpiece connected to the breathing circuit. The rebreathing method was used in order to provoke graded hypercapnia. Steady-state end-tidal (partial) carbon dioxide pressure (PET_CO2) values were obtained by adding a combination of oxygen and a binary mixture of 7.0% CO₂ in O₂. The total resistance of the circuit was 0.32 cm/L/s at 1 L/s flow. A visual display on screen was used to target ventilation according to the experimental protocol.

Borg Scale

Before rebreathing started, subjects were instructed to use the verbal descriptors of the modified Borg scale (12) to quantitate the sense of inspiratory effort and report the associated number. Intermediate values between verbal descriptors were accepted. Instructions were given to the subjects to include only respiratory sensations.

Measurement Variables

Inspiratory and expiratory air flow was measured with a heated pneumotachograph (Fleisch No. 4) attached to the mouthpiece and connected to a pressure transducer (Validyne MP 45, Northridge, CA). PET_CO2 was measured at the mouth using an infrared technique (Oscar-Oxi; Datex, Helsinki, Finland). The CO₂ analyzer was calibrated with room air and a gas mixture containing 5.0% CO₂ and 54.5% O₂ (Quick Cal; Datex), prior to every study. PET_CO2 was used as an indirect estimate of arterial PCO₂. Oxygen saturation was measured with a pulse oximeter placed on a finger (Oscar-Oxi; Datex).

Signal Processing

Flow signal and PET_CO2 were simultaneously displayed on a four-channel polygraph pen recorder (Physiograph MK-IV-P; Narco Bio-systems, Houston, TX) and stored in a magnetic tape PCM recording adapter (Vetter digital model 4000 A; Vetter, Rebersburg, PA). For acquisition, the signals were passed through a 16-bit analog-to-digital conversion board (Sponge Inc. Data Translation, Sarasota, FL) at a sampling rate of 60 Hz and stored in a computer (IBM PS/2) for future off-line analysis. The signal analysis was carried out breath by breath with Start IS software (Start, Montreal, PQ, Canada). Tidal volume was obtained via integration of the flow signal into the computer.

Experimental Protocol

The protocol included five steps: (1) basal measurements after a stabilization period with the subject connected to a mouthpiece (BASAL); (2) rebreathing technique to provoke hypercapnic hyperventilation (HV); (3) screen copy of tidal volume and respiratory frequency (COPY) at the same hypercapnic level (possible distraction effect); (4) maintaining the same screen copy at basal PET_CO2 values (ISO); and (5) room air breathing with untutored spontaneous ventilation (REC). Each step duration was 5 min. Borg readings were demanded every 30 s. Values of the last 2 min were reported.

Statistical Analysis

Analysis of variance was used on data obtained in the five steps. If differences were found, we compared paired data. Significance for this test was considered to be p < 0.01. Data in the text and tables are expressed as means ± SEM.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The rebreathing procedure increased PET_CO2 from a basal value of 40.6 ± 0.44 to 54.0 ± 0.81 mm Hg (p < 0.001) and ventilation (VE) from 12.0 ± 0.50 to 28.1 ± 1.19 L/min (p < 0.001). The Borg value increased from 0.11 ± 0.06 to 3.4 ± 0.23 (p < 0.001) (Figure 1). The values during HV and COPY periods were not statistically different.

View larger version (16K): [in this window] [in a new window]   Figure 1.   PETCO2, VE, and Borg rating during five steps of protocol: BASAL = basal measurements; HV = hypercapnic hyperventilation, COPY = screen copy of ventilatory pattern; ISO = the same screen copy at basal PETCO2 values; REC = recovery.

During the ISO period PET_CO2 was 40.2 ± 0.59 L/min (not significant [NS] compared with basal), while VE was 29.9 ± 1.29 L/min (NS compared with HV and COPY). On the other hand, the Borg value during the ISO period decreased to 1.86 ± 0.28 (p < 0.001 compared with HV and COPY). All REC values were not different from the BASAL condition (Figure 1).

Variations in the IES during the HV period were entirely dependent on PCO₂. Isocapnia resulted in IES decrease during similar hyperventilation levels (Figure 2). The slopes of IES/ VE during the HV, COPY, and ISO periods were 0.20, 0.17, and 0.09, respectively.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Ventilation-Borg score relationship during HV, COPY, and ISO periods. Numbers in the figure represent the PETCO2 values during each step.

The respiratory frequency (rf), tidal volume (VT), inspiratory time (TI), duty cycle (TI/Ttot), and mean inspiratory flow (VT/TI) are shown in Table 1. Recovery values were not statistically different from the basal values. The HV, COPY, and ISO values did not change significantly.

The threshold PCO₂ and VE for the first change in the Borg score from the basal value are shown in Figure 3. A wide range of PCO₂ and VE values was observed. In three subjects Borg values increased before changes in the ventilation (Figure 3B). The time course of PCO₂ increase was from 0.024 to 0.264 mm Hg/s. Subjects with a higher rate of PCO₂ increase noticed the first change of IES at higher PCO₂. The PCO₂/time relationship was positively correlated (r = 0.70, slope = 0.099, p < 0.001).

View larger version (20K): [in this window] [in a new window]   Figure 3.   Basal PETCO2 (A) and ventilation (B) are plotted versus PETCO2 and ventilation at the moment of the first change in the breathing sensation (threshold PCO2 and ventilation).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The increased IES induced by hypercapnic hyperventilation was significantly and considerably reduced when the same hyperventilation level was voluntarily maintained under isocapnic conditions. We suggest that CO₂ per se generates an IES independent of the concomitant increase in ventilation. Before accepting this statement, several possibilities should be considered.

In order to assess the role of CO₂ upon IES or the urge to breathe, several approaches have been used: breath-holding technique in normal subjects (2, 3, 11), exercise plus added CO₂ (4), hypercapnic hyperventilation (9), chronically ventilated patients (5, 6), and conscious, paralyzed volunteers (7, 11).

Oscillating and continuous CO₂ administration have been used to assess the CO₂ effects upon IES. Adams and colleagues (13), using an oscillating pattern of inhaled CO₂, assumed that the sensation assessed by visual analogic scale follows ventilation and not CO₂. However, both stimuli may have different latency times to provoke changes in respiratory sensations. Thus, it is difficult to know if there was an overlapping in latency time for CO₂-associated sensations and ventilation. This critique is only valid assuming a separate effect of CO₂ and ventilation upon IES. On the other hand, Demediuk and coworkers (14), using a steady-state protocol of hypercapnia imposed upon isocapnic hyperventilation, suggest that effort sensation can be differentiated from breathlessness. However, it is not easy to accept that the verbal descriptors used represent different physiologic mechanisms.

In the present study, the rebreathing method was used to provoke graded hypercapnia. The rate of PET_CO2 increase in our subjects was variable; when the subjects noticed a change in sensation, then the PET_CO2 was maintained relatively stable. The reported Borg value was selected after 2 min of stable PCO₂. With our protocol, we assume that the IES potentially reflects the participation of both increased ventilation and hypercapnia, i.e., both stimuli during HV and only ventilation during the ISO period.

Changes in ventilation or CO₂ level could be associated with different respiratory sensations. Since hypercapnia induces hyperventilation, it is important to dissociate breathing discomfort induced by both stimuli. Patients with their complete high-cervical spinal cord section chronically ventilated develop "air hunger" during CO₂ administration, despite the absence of chest wall afferent information (6). Similarly, in paralyzed normal volunteers, CO₂ induced the sensation of air hunger in the absence of voluntary muscle contraction (7).

Since the spontaneous pattern of breathing minimizes respiratory sensations, we allowed a free breathing pattern during HV, which was copied during the ISO period. Since the VE and ventilatory pattern and VT/TI were unchanged during the HV, COPY, and ISO periods (Figure 1, Table 1), we ruled out possible mechanical influences upon changes in the IES.

It is interesting to analyze VE and PET_CO2 at the moment of the first change in the IES. The PET_CO2 and VE level at this moment were variable (Figure 3). In three cases changes in respiratory sensation preceded changes in VE (Figure 3B). Possibly on account of the different individually adopted ventilatory patterns, the time course of PCO₂ was variable. Subjects with higher rates of PCO₂ increase noticed the first change of IES at higher PCO₂. However, since PCO₂/time relationship was positively correlated, differences in the IES detection in the setting of the broad PCO₂ range detection suggest individual differences in the PCO₂ sensitivity.

Freedman and colleagues (10) studied patients with chronic airflow limitation during CO₂ rebreathing and voluntary isocapnic copy of their previous breathing pattern. Voluntary copy resulted in less breathlessness. They suggested that the sensation of breathlessness is more dependent on the awareness of central processing than on input from peripheral mechanoreceptors.

Chonan and coworkers (9) compared the sensation of difficulty in breathing during progressive hypercapnia and during isocapnic voluntary hyperventilation in normal subjects using a modified Borg scale. Hypercapnic hyperventilation produces more IES than voluntary isocapnic hyperventilation. This fact was attributed by the authors to a different neurologic pathway. However, in both studies restoring a normal PCO₂ could be the reason of decreased IES.

We conclude that under our experimental conditions of steady-state CO₂ rebreathing in normal subjects, CO₂ produces an IES, independent of the ventilatory changes induced. The threshold response to CO₂ in terms of the first change in IES was variable; individual differences may play some role.