| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of 1Physiology and 2Anaesthesia, University of Toronto, Toronto, Ontario, Canada M5S 1A8
Submitted 19 March 2004 ; accepted in final form 21 July 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
exercise hyperpnea; wakefulness drive; afferent feedback
E) increases immediately (25), and the rapid onset of this drive to breathe has led to it being assigned a neural origin, since it is too fast for a humoral source (15, 26). This "exercise drive to breathing" is hypothesized to result from two sources: one is "peripheral neurogenic drive," whereby afferent feedback from the exercising limbs leads to increased respiration (21); the other is "central command," whereby motor commands to the limbs initiate a parallel activation of respiration (17, 49). In the integrated system, both of these mechanisms are initiated simultaneously; therefore, any interaction between them could affect their respective contributions toward the control of breathing. Indeed, animal model studies have provided evidence of interactions between central motor commands and afferent feedback (2, 13, 37). Such findings support the idea that, at the onset of exercise, neural drives to breathe via distinct mechanisms are incorporated to provide the drive to breathe that is observed in the integrated system.
The onset of exercise also involves an increase in central neural activity via cortical brain activity (23, 32) that is not solely related to the motor outflow to the muscles itself. For example, during imagined exercise where both central motor command and afferent feedback mechanisms are absent, significant increases in cortical activity occur and this activity is able to increase breathing (47). The effect of this brain activity or arousal, which is independent of actual motor command, on the afferent feedback mechanism is presently unknown.
We do know that forms of neural arousal apart from exercise affect respiratory control in humans (34, 40). Such arousal-related influences are ambient light and sound (19, 44), and cognitive tasks, such as solving puzzles and mathematical problems (24, 41, 43). A withdrawal of such central neural arousal occurs in sleep (33, 40, 42, 48), and the drive to breathe that is related to this arousal or behavior is therefore also withdrawn, leading to it being referred to as a "wakefulness drive to breathe" (18). Interestingly, a reduction in arousal can also be effected through meditation, and the resulting decreases in respiratory rate and tidal volume (VT) (46, 51) demonstrate that the wakefulness drive to breathe is also affected. We therefore use the term "wakefulness" to describe a central neural phenomenon that is similar to alertness or awareness, attributing it to increased central neural activity or arousal.
Given the evidence suggesting there is interaction between central motor activity and afferent feedback in animal preparations, we wished to investigate whether central neural activity related to the state of wakefulness would also affect the afferent feedback mechanism in humans. To increase wakefulness in our experimental subjects, we used a cognitive task, solving a puzzle; to stimulate the afferent feedback mechanism, we used passive leg extension movements. With passive movement of a subject's limbs in a rhythmic, exercise-like fashion, mechanical changes in the skeletal muscles can generate afferent feedback and thereby increase E (5). Because central motor command is absent during this process, the transition from rest to passive limb movement allows the discrete study of afferent feedback and how its contribution to respiratory control may be affected by increased wakefulness via a cognitive task.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Subjects were asked to visit the laboratory on two occasions. During the first visit, a familiarization session, subjects were shown the equipment and techniques that we would use, introduced to the sensation of passive leg movements in the tandem chair apparatus, and familiarized with a computer-based puzzle called "The Tower of Hanoi" (Fig. 1). A second visit, no more than 1 wk later, was arranged to perform the experimental protocol in which the subjects completed two passive leg movement sessions, identical except that during one they relaxed and during the other they solved the computer-based puzzle.
|
70° (270° at maximum flexion to 200° at maximum extension). E, VT, breathing frequency (fB), heart rate, mean arterial blood pressure (MAP), end-tidal PCO2 (PETCO2), mixed expired PCO2 and PO2, leg movement frequency (fLM), and leg muscle activity (EMG) were monitored noninvasively throughout all procedures. Subjects were comfortably seated in the chair apparatus and were monitored for 3 min at rest before beginning 3 min of passive leg extensions. Monitoring continued through another 3 min of rest after the cessation of passive movement. This common procedure was performed twice, under the following two conditions. For condition 1, the subject is awake, seated in the chair, and wearing a blindfold and ear protectors ("ear-muff" type, –29 decibels) to minimize arousal via visual and auditory stimuli, throughout this entire data collection interval. This condition will be referred to as the low wakefulness (LW) protocol.
For condition 2, the subject is awake and is seated in the tandem chair while solving a computer-based puzzle called the "Tower of Hanoi" on a notebook computer to increase arousal throughout this entire data collection interval. This condition will be referred to as the high wakefulness (HW) protocol.
Computer-based puzzle.
In the HW protocol, subjects were solving a puzzle called the Tower of Hanoi, shown schematically in Fig. 1. Subjects controlled the game using a mouse connected to a laptop computer placed at eye level 4 ft in front of them, while they sat in the tandem chair. The mouse was placed on the right-hand side of the chair on a mouse pad that was fixed to the frame of the chair. The specific program was, at the time of this study, available as freeware on the internet (http://www.mazeworks.com/hanoi/, Copyright 2002 David Herzog; based on the puzzle invented by French mathematician Edouard Lucas in 1883).
Equipment. Subjects breathed through an air-cushioned facemask (model KM201/202/203, Vacu·Med, Ventura, CA), and breathing was monitored with a turbine (universal ventilation meter, model 17125, Vacu·Med). The turbine was connected to a two-way non-rebreathing valve (model 2630, Hans Rudolph, Kansas City, MO), enabling the subject to inspire room air while their expirate was collected in a 7-liter compliant mixing bag that passively emptied into the room. We monitored PETCO2 using an anesthetic gas monitor (type 1304, Bruel and Kajer) that sampled via a port on the inspired side of the turbine. Mixed expired PO2 and PCO2 were monitored via gas analyzers (models CD-3A and S-3A/1, respectively, Applied Electrochemistry, Pittsburg, PA) sampling via a catheter within the mixing bag. We monitored heart rate and MAP on a beat-to-beat basis using a Portapres (model 2.0, FMS, Arnhem, The Netherlands). Root-mean-squared leg muscle EMG was monitored via surface electrodes (Meditrace 133, Ludlow LP, Chicopee, MA) placed over the vastus lateralis of the left leg, using the suprailiac region of the lateral abdomen as a ground site. The raw EMG signal was passed through a two-stage preamplifier (Neurolog model NL 100 and NL 104 in series, Digitimer, Hertfordshire, UK) and a custom-built downstream AC filter (high pass >10 Hz, low pass <500 Hz).
The experimenter controlled a data-acquisition computer (Inspiron 8100, Dell) from a seated position in the rear seat of the tandem chair, interfaced via a mouse, and maintained the pace of limb movement (fLM) by following a visual metronome integrated into the data-acquisition software. fLM was monitored with a mercury switch mounted to the left extension arm of the chair and connected to a digital input on the data-acquisition system. The continuous analog output from all monitoring devices was also passed through a pulse-code modulation recording adapter (model 4000A, A. R. Vetter) to allow archiving of data on VHS tape (JVC Hi-Fi Stereo VCR, model HR-D840U 500C, JVC Americas). Analog signals were fed to a 16-bit data-acquisition card (DAQCard-AI-16XE-50, National Instruments, Austin, TX) installed in the data-acquisition computer. Data-acquisition software was custom designed (LabVIEW 6.1, National Instruments; source code available on request). Signals were displayed in real time (sampling rate = 20 Hz), and data were analyzed and collected on a breath-by-breath basis and written to a text array file for further analysis. The exception to this breath-by-breath analysis of data was the collection of metabolic gas-exchange data, which was performed by the software on the basis of mixed-expired values as averaged over intervals corresponding to the collection of 4 liters of expired air in the mixing bag.
Data analysis.
Data collected during the experiment were analyzed with a specially written analysis software (Labview 6.1, National Instruments). This software uses a line-fitting technique that minimizes the mean standard error to determine an equation that best describes the data over each of the three discrete intervals of the test: 1) rest, 2) passive limb movement, and 3) recovery. Data collection began for a subject once all monitored parameters appeared to be stable; thus the resting parameters (E, VT, etc.) were fitted with a straight line of zero slope. The software then used the Levenberg-Marquardt algorithm to fit an exponential rise or decline or a dual-exponential rise or decline to the data of the passive movement and recovery. The equation that better described the data (smaller mean squared error) was chosen as the representative equation. From these line equations, data points were then calculated for times corresponding to the start of the phase, and each 15 s thereafter, up to and including the 3-min mark (i.e., 180 s) for that interval. In other words, data points were determined along the regression lines at times of 0, 15, 30, 45, ..., 180 s for the first rest interval. For the interval of passive limb movement, data points were determined at times of 180, 195, 210, ..., 360 s. For the 3-min interval after the cessation of passive limb movement, data points were calculated at times corresponding to times of 360, 375, 390, ..., 540 s.
To determine whether our method of altering wakefulness drive to breathe was effective, we compared resting (premovement) values for E, VT, fB, PETCO2, heart rate, MAP, and metabolic gas exchange [oxygen uptake (O2) and carbon dioxide production (CO2)] between the two conditions (HW vs. LW). Subjects were self-matched for comparison; therefore, all statistical testing involved repeated measures (RM) analysis. We used one-way RM ANOVA testing and Bonferroni's post hoc analysis to test for any significant effects of the background cognitive state on the resting measures. Tests were conducted using an a priori level of significance (
) set at 0.05.
To examine the behavior of observed parameters during passive movement in the two background conditions, we used two-way RM ANOVA, again with Bonferroni's post hoc testing. Experimental condition (LW and HW) was one factor considered in analysis, with limb movement status (rest and passive movement) as a second factor. Again, all analyses were performed with an a priori level of significance () set at 0.05. Analysis was performed with the binned data calculated from the fitted equations at 15-s intervals. E, VT, fB, heart rate, and MAP were analyzed for evidence of fast changes at the onset and cessation of passive movements by performing a two-way RM ANOVA on four data points corresponding to 1) the end of the premovement rest interval, 2) the beginning of the passive movement interval, 3) the end of the passive movement interval, and 4) the beginning of the postmovement rest interval. Significant changes in these parameters at the onset and cessation of passive movements were interpreted to be an index of the drive to breathing provided by peripheral neurogenic drive.
Leg movement pace, fLM, and leg muscle activity (EMG) for the two background conditions (LW and HW) were compared between states (rest and passive movement) with a two-way RM ANOVA, using data points corresponding to the last point of rest before passive movement and the last point of passive movement. PETCO2, O2, and CO2 were analyzed in the same manner, enabling a comparison between parameters at rest and values closest to what would be considered "steady-state" passive movement, in both LW and HW conditions.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
E, fB, and VT data for one subject in both LW and HW conditions are shown in Figs. 2 and 3. Figure 4 shows data for respiratory measures averaged across all nine subjects in both LW and HW conditions.
|
|
|
E, VT, and fB averaged across all subjects in both HW and LW conditions at the four points of interest in the protocol are shown in Table 1. Mean resting E and fB were higher during HW than during LW (P < 0.001 and P < 0.004, respectively). There were no differences in resting VT between LW and HW conditions (P = 1.00). All subjects exhibited greater E at rest during HW compared with LW (Fig. 4).
|
E (P < 0.001), fB (P = 0.001), and VT (P = 0.017), demonstrating that all three respiratory parameters increased at the start of passive movement (see Figs. 5 and 6). There was an interaction effect between the factors of movement status and the wakefulness condition present for the three respiratory measurements of E (P = 0.002), fB (P = 0.004), and VT (P = 0.002). In other words, background wakefulness condition affected the size of the fast changes in E, fB, and VT that resulted from the passive leg movement. In the LW condition, fast increases were observed in all three parameters at the onset of passive leg movement (see Table 2 and Fig. 6). In the HW condition, only fB increased significantly, but this increase was smaller than the corresponding increase in the LW condition.
|
|
|
E (P = 0.011) and fB ( P < 0.001) but not in VT (P = 0.53). In the HW condition, neither E (P = 1.000) nor its components fB (P = 0.074) and VT (P = 1.000) demonstrated significant fast changes at the end of passive leg movement. PETCO2. The behavior of PETCO2 is illustrated in Fig. 5D. Resting PETCO2 was significantly lower (P = 0.002) during HW (36.8 ± 0.6 Torr) than during LW (38.6 ± 0.8 Torr). There was neither evidence of significant treatment effects for the factors of wakefulness condition or movement status nor evidence of an interaction between these factors.
Cardiovascular measurements. Average heart rate and MAP for all subjects during both HW and LW conditions at the four points of interest in the protocol are shown in Table 1. Resting heart rate was significantly higher in HW than during LW (P = 0.017); no such differences were present in resting MAP (P = 0.860). Passive leg movements had a significant effect on heart rate (P < 0.001), but the fast changes at the start and end of passive movement were not significant (P = 0.197 and P = 1.00, respectively). Rather, a significant difference was present between heart rate at the start and end of the 3 min of passive leg movement (P < 0.001). Passive movements had no significant effect on MAP (P = 0.763). There was no effect of wakefulness condition on either heart rate (P = 0.111) or MAP (P = 0.994). In addition, there was no evidence of interaction effect between passive movement status and wakefulness condition for heart rate (P = 0.205) or MAP (P = 0.910)
Metabolic gas exchange.
Metabolic gas-exchange data are summarized in Fig. 7. There were no significant differences at rest during LW and HW in terms of either O2 (P = 0.214) or CO2 (P = 0.190). Neither wakefulness condition nor passive movement status significantly affected O2 (P = 0.288 and P = 0.724, respectively) or CO2 (P = 0.529 and P = 0.141, respectively), and these factors did not interact (P = 0.994 and P = 0.471, respectively).
|
Leg muscle activity. The EMG signal changed significantly during passive movement, independent of wakefulness condition (P = 0.024). There was no independent effect of wakefulness condition (P = 0.659), and no interaction effect was present between wakefulness condition and movement status (P = 0.436). The average EMG activity at rest for both conditions combined was 1.6 ± 0.2 µV, which increased to 2.1 ± 0.3 µV during passive movement.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
E, resulting from an increase in fB, compared with the LW condition. The start of passive leg movements in the LW condition caused a fast increase in E, fB, and VT. In the HW condition, the respiratory response to the onset of passive leg movement was different; there was no increase in either E or VT, and the increase in fB was significantly smaller. There were no apparent effects of passive leg movements on the cardiovascular measurements of heart rate and MAP, apart from a decline in heart rate during the 3 min of passive limb movement in either condition. Wakefulness condition. In the LW condition, subjects were blindfolded and wore ear protectors that minimized auditory and visual stimuli. In the HW condition, subjects solved a computer-based puzzle game with minimal incentive to quickly move the hand or fingers so as to minimize the use of musculature during the protocol; indeed metabolic gas exchange at rest was not significantly affected by the wakefulness condition.
In previous studies reporting the effect of wakefulness on resting E, VT appeared to be less affected than fB (3, 10, 41, 44), and our findings agree; E was 20% greater during rest during HW than during LW, and this was achieved through an elevated fB with VT unaffected. The higher resting E during HW compared with LW was sufficient to cause significantly lower PETCO2 levels. These findings are consistent with the present knowledge of the effect of wakefulness on the control of breathing at rest; therefore, we are confident that we provided subjects with two conditions with the desired difference in background wakefulness drive.
The additional central stimulation of E in the HW condition caused a mild hyperventilation at rest, lowering PETCO2 by 2 Torr compared with the LW condition. Although such a difference between conditions provided a mechanism that may have altered the ventilatory response at the start of passive leg movements, the literature suggests otherwise. There are three studies that suggest that hypocapnia may reduce or eliminate the fast increase in E at the start of exercise attributed to fast neural drives to breathe (1, 22, 50). However, the present consensus in literature suggests that the decrease in PETCO2 that we observed in the HW condition at rest was unlikely to affect the fast changes in E observed at the onset of limb movement (8, 9, 16, 28–30). These studies collectively show that mild to moderate acute hypocapnia does not affect the fast neural drives to breathe at the start of exercise. A thorough discussion of the issues surrounding this topic appears in the most recent of these papers (27). We are therefore confident that the differences that we observed between the LW and HW conditions cannot be attributed to a difference of 1.8 Torr in PETCO2 before passive movements.
Effectiveness of passive leg movement.
We used a method of passive movement that we have previously described and validated (4, 5). The onset of passive leg movement is likely to cause a return of pooled venous blood from the lower limbs into the central circulation that will transiently alter respiratory control. However, we were interested in the fast (nearly immediate) changes in E at the onset of passive limb movement. Such fast changes occur before the pooled venous blood alters central arterial PO2 and PCO2, thereby stimulating E. Therefore, although passive leg extension movements do not isolate afferent feedback from all other respiratory drives throughout the movement period, they do isolate afferent feedback from central motor drive during the rapid phase of ventilatory adjustment at the start of limb movement.
Passive leg extensions were effective, increasing E over resting values. Because these changes were observed at the onset of passive leg movement, we attribute them to peripheral neurogenic drive due to stimulation of mechanically sensitive, lightly myelinated or nonmyelinated afferent fibers in skeletal muscles (21).
That the leg movements were passive is attested to by the small increase in EMG activity. Although EMG activity was significantly increased by 1 µV on average, this was far less than the 20–50 µV that we routinely observe during light to moderate active exercise of the same modality and frequency. We therefore attribute the observed increase to movement artifact and not to an active exercise component. Further corroboration of the passive nature of the movement was the lack of increase in O2 and CO2.
Interaction between peripheral neurogenic drive and wakefulness. These experiments provide the first evidence in awake humans that the ventilatory response to the afferent feedback mechanism is altered by the level of cognitive activity. Ishida et al. (20) measured the ventilatory response to passive leg movements during sleep and awake states in human subjects and found that the fast changes were greater during sleep; however, these differences were not statistically significant. Nonetheless, their study does provide corroborative support for our observations. Our study only differs in that our subjects were awake in both conditions; we varied the level of arousal or wakefulness via a cognitive task, and significant changes in respiratory control were observed.
We have also shown that the background wakefulness condition affected the behavior of both the VT and fB components of E during passive movement; there were fast changes evident in all three measures during LW. By contrast, in the HW condition, fast changes at the onset of passive movement were only evident in fB and absent in both VT and E.
Mechanism and site of interaction.
Whereas our results suggest that the wakefulness associated with a cognitive task suppresses the afferent feedback mechanism that increases E during passive movement, the nature and location of this interaction remain speculative. Increased wakefulness appears to produce two independent changes in respiratory control: it both augments E at rest and suppresses the ventilatory response to afferent feedback. There is therefore likely a site of interaction outside the respiratory rhythm generation network because inhibition or disfacilitation of the respiratory rhythm generation network itself would result in a decrease in E at rest.
A cognitive task has been shown to affect the H reflex in humans (38). It is therefore possible that neural activity via a cognitive task may modulate other neural signals at the level of the spinal cord in a manner similar to the way in which central motor command that has been shown to interact with afferent feedback (14). Alternatively, the interaction could take place in higher regions of the central nervous system, which would agree with the view of Ishida et al. (20). Sites that influence respiratory control that are known to receive sensory input and/or respond to states of arousal include medullary reticular neurons (6, 11, 45), the retrotrapezoid nucleus (7, 12, 31), the nucleus reticularis gigantocellularis (37), and the nucleus tractus solitarius (35, 36, 39).
Potential significance for exercise hyperpnea. If the respiratory response to the afferent feedback of passive limb movement is suppressed when wakefulness is increased, is afferent feedback an effective drive to breathing during active exercise in humans? During active exercise, there is central nervous system activity that is not directly related to the generation of central motor command but rather is concerned with the planning of movement, goal-oriented behavior, anticipation of events, and other thought processes. Therefore, there is likely to be a significant increase in cognitive activity independent of the central command mechanism at the start of active exercise. We therefore speculate that such increased arousal or wakefulness during exercise may effectively suppress the peripheral neurogenic drive to breathe from afferent feedback, rendering that mechanism ineffective during active exercise in the integrated system.
Nevertheless, we emphasize the following caveats. First, the wakefulness associated with exercise may be quite different from that associated with solving a puzzle, and thus the respiratory consequences may be equally different. Second, there may be differences in the quantitative and/or qualitative nature of the afferent feedback and how it interacts with wakefulness drives during normal exercise.
Conclusion. Increased wakefulness via a cognitive task suppresses the respiratory response to passive limb movement. However, the relevance of this finding toward the control of exercise hyperpnea will require future investigation into the nature and behaviors of central neural arousal and afferent feedback during exercise.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. J. Bell, W. Feenstra, and J. Duffin The initial phase of exercise hyperpnoea in humans is depressed during a cognitive task Exp Physiol, May 1, 2005; 90(3): 357 - 365. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and HW (
) conditions. A: 
significant between HW and LW groups, P 
0.004; **significant between rest vs. the start of passive leg movement, P 