INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

DYNAMIC EXERCISE increases heart rate, blood pressure, and ventilation, and the mechanisms involved include feed-forward ("central command") and feedback mechanisms. Central nervous influence on the increase of heart rate and ventilation is evident in anticipation of exercise (24), during hypnotic suggestion of intense exercise (18), and even during imagination of exercise (26). Central command may be of importance for the heart rate response to exercise with a small muscle mass, but it cannot account entirely for the heart rate response to exercise with large muscle groups (14, 21). Neural feedback from the exercising legs is important for the blood pressure response to cycling because this response is attenuated by epidural anesthesia (7) or abolished, depending on the intensity of the block (4, 5, 8, 12, 22). However, epidural anesthesia or spinal cord injury (13) is of little importance for the elevations in heart rate and ventilation during electrically induced exercise. The tight coupling between the exercise cardiac output and the oxygen uptake, even in the absence of intact neural feedback and motor center activity, indicates potential importance of blood-borne factors ("humoral feedback") for the exercise heart rate. Thus, during dynamic exercise, tachycardia could be influenced by metabolites arising from the working muscles, by translocation of blood to the heart, and/or by an elevated blood temperature (12, 15, 20).

We hypothesized that the exercise elevation of heart rate would be reduced if blood-borne feedback from the working muscles is absent. Therefore, we compared the response to electrically induced leg exercise (3) both with and without pneumatic cuffs around the thighs in persons with spinal cord injury with that elicited during voluntary cycling in able-bodied subjects. Sensory block of cardiovascular importance was evidenced by the recording of the arterial blood pressure (12, 22).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Three persons with paraplegia and four with tetraplegia [1 woman, 6 men, age 36 ± 7 (SD) yr, weight 79 ± 5 kg, and height 180 ± 7 cm] together with seven able-bodied volunteers, matched by gender and age (age 32 ± 4 yr, weight 77 ± 13 kg, and height 186 ± 8 cm), participated in the study after informed consent to the protocol, which was approved by the Ethics Committee of Copenhagen (KF-W92-061). The level of the spinal cord injury was at the fourth thoracic vertebra in the participants with paraplegia and at the sixth cervical vertebra in the participants with tetraplegia. The time since injury ranged between 2 and 23 yr (mean 11 yr). A clinically complete motor block affecting the lower extremities was present in all participants with spinal cord injury, and five participants had also a complete sensory block (Frankel class A). Two of the participants with tetraplegia had some sensory function retained (Frankel class B). The participants with spinal cord injury had participated in a 1-yr program with training three times per week by using the same ergometer as applied in this study.

Electrically induced cycling. The spinal cord-injured subjects were placed in the sitting position, fastened with a belt around the hips, on a computer-controlled ergometer (REGYS I Clinical Rehabilitation System, Therapeutic Alliance, Dayton, OH). This system is composed of three subsystems: a lower extremity cycle ergometer (Monark 4000, Stockholm, Sweden), a stimulus control unit, and a reclinable patient chair. A computer controls and monitors the electrical stimulation according to prescribed parameters. Surface electrodes (2 × 4 cm) were placed over the motoneuron end plates (motor points) where the stimulation threshold was lowest for the quadriceps, hamstrings, and gluteal muscles of both legs. Two electrodes were applied over each muscle group coated with a buffered electrode gel. Six channels for sequential surface muscle stimulation were used with a computer-controlled closed-loop system. Each channel supplies monophasic rectangular pulses lasting 350 ms delivered at 30 Hz. Stimulation intensities ranged from the preset threshold determined for each muscle group to elicit a palpable contraction (range 18-40 mA) to a maximum of 130 mA. A pedal-position sensor allowed for calculation of velocity and was also used to control the instantaneous stimulus amplitude required for each of the six muscle groups to result in a smooth cranking frequency of 50 rpm. Stimulation started with a 2-min period of 50% threshold intensity and then continued with the full stimulation.

Procedures. After the subjects rested for 10 min, one of the investigators moved the pedals for 2 min, and this was followed by electrically induced cycling with no external load ("0 W") on the ergometer for 15 min and then for 15 min at a maximal load that it was possible to maintain over the work period (8 W, range 6-12 W). This workload was known to elicit oxygen uptake rates of ~1 l/min in similar persons with spinal cord injury (13). The able-bodied participants were actively cycling with no external load on the ergometer for 15 min followed by 15 min at a work rate of 8 W and then for 15 min aiming at the average oxygen uptake (~1 l/min), the same workload accomplished by the participants with spinal cord injury during loaded exercise.

Cardiorespiratory measurement. Heart rate was calculated as the inverse of the beat-to-beat interval of the middle cerebral artery blood velocity tracing (FAST-system vs. 2.1, TNO Institute of Applied Physics, Amsterdam, The Netherlands). During cycling with free circulation to the legs, all participants with spinal cord injury had blood pressure measured from a catheter in the femoral artery. During the exercise with vascular occlusion, two participants with spinal cord injury had finger arterial pressure measured (Finapres, Ohmeda, Madison, WI) (see Table 2). The determination with this method yielded blood pressure readings similar to blood pressure measured in the radial artery of this population (unpublished observations). In the able-bodied participants during cycling with free circulation to the legs, blood pressure was determined by a sphygmomanometer (Tycos, Taylor Instrument, Asheville, NC). Ventilation, breathing frequency, carbon dioxide elimination, end-tidal carbon dioxide tension, and oxygen uptake were measured breath-by-breath (CPX/D, MedGraphics, St. Paul, MN).

Statistical analysis. Values are means ± SE for the average of the beat-to-beat or breath-by-breath measurements over the last 2 min of the resting phase before exercise and for the following 3 min in both voluntary and electrically stimulated exercise. In the passive cycling period before electrical stimulation and during vascular occlusion at rest and after exercise, values are derived from averages over 2 min. The Friedman test was used to determine whether significant changes took place with time or between circumstances, and such changes were located with the Wilcoxon signed-rank test. The Mann-Whitney U-test was used for comparison between groups. A P value of <0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Rest. In the able-bodied individuals, heart rate was 66 ± 4 beats/min and similar to that of the participants with spinal cord injury (73 ± 3 beats/min). Also, ventilatory variables were not significantly different between the two groups, with the exception of a lower oxygen uptake in the spinal cord-injured subjects that reached a statistically significant level before the thigh cuffs were inflated (0.30 ± 0.05 vs. 0.36 ± 0.05 l/min; Fig. 1). Mean arterial pressure was lower in the tetraplegic compared with control subjects (64 ± 4 vs. 88 ± 3 mmHg; see Table 2).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Heart rate (HR) and ventilatory variables during cycling with free circulation to the legs. HR, ventilation (E), breathing frequency (f), carbon dioxide elimination (CO2), and oxygen uptake (O2) during electrically induced cycling in spinal cord-injured participants (, ) and in able-bodied subjects (, ) during voluntary cycling at the same absolute work rates, at a O2 of ~1 l/min, and during the first 2 min of recovery (R1, R2). Values are means ± SE. Vertical dashed line, end of time interval taken to compare values with vascular occlusion exercise in spinal cord-injured subjects; vertical dotted lines, time points at which a transition between rest, exercise, or cuff is undertaken. For each studied parameter, filled symbols are values significantly different from resting values (P < 0.05). * Significant differences between experimental groups at specific time points, P < 0.05.

Exercise. During passive movement of the legs, heart rate remained statistically unchanged in both studied groups (Fig. 1) but ranged from a drop of 25 beats/min in a tetraplegic individual to an increase by 15 beats/min in one paraplegic subject (Table 1). Oxygen uptake and carbon dioxide elimination increased in the able-bodied subjects, but they did not change in the spinal cord-injured subjects.

Exercise with thigh cuffs. In contrast to exercise with free circulation to the legs, ischemic cycling elevated ventilatory variables to a similar extent in both groups of subjects (Fig. 2). For the able-bodied individuals, the increase in ventilation was lower during ischemic cycling (20 ± 2 l/min) than during free-circulation cycling (25 ± 1 l/min). During postexercise muscle ischemia, ventilatory variables remained above the resting level in the para- and tetraplegic subjects, whereas they recovered in the able-bodied individuals. In one participant with paraplegia, mean arterial blood pressure increased immediately after inflation of the thigh cuffs and remained elevated until deflation, whereas, in one participant with tetraplegia, mean blood pressure continued to increase during ischemic exercise (Table 2).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   HR and ventilatory variables during cycling with vascular occlusion of the legs. Workload for spinal cord-injured subjects was 0 W and for able-bodied subjects corresponded to load established during highest exercise load with free circulation to the legs. Values are means ± SE. Figure description is the same as for Fig. 1. For each studied parameter, filled symbols are values significantly different from resting values (P < 0.05). * Significant differences between experimental groups at specific time points, P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study presents data from dynamic leg exercise in humans in whom the increase in heart rate was reduced or even eliminated by arterial occlusion of the legs. Previously, epidural anesthesia has been shown to blunt the blood pressure but not the heart rate response to dynamic exercise (12, 22). Similarly, in this study there was no significant increase in blood pressure during cycling in participants with spinal cord injury, supporting the idea that peripheral neural signals dominate the regulation of blood pressure but not that of heart rate.

We made the assumption that neural feedback from the exercising muscle was absent in the individuals with spinal cord injury. This assumption was supported by the finding of a normal blood pressure response to exercise that was blunted by epidural anesthesia (12, 22) and was also absent in the participants with spinal cord injury. The cardiovascular responses of the spinal cord-injured individuals to electrically stimulated exercise are in accordance with those observed by others (6). In contrast, Brice et al. (4) found that blood pressure was higher in untrained spinal cord-injured individuals compared with controls, probably as a result of autonomic hyperreflexia, and the absence of any heart rate response in that study was taken to reflect stimulation of the baroreflex.

For comparison between the two groups of participants during exercise with free circulation to the legs, the very low workload applied has to be considered. Even at a workload of 0 W, the spinal cord-injured individuals increased their heart rate markedly toward the end of that exercise bout (Fig. 1), and at a maximal work rate (8 W) their pulmonary oxygen uptake was only ~1 l/min. In the able-bodied individuals, who performed volitional cycling standardized either to the same absolute external load or oxygen uptake, heart rate increased less than during evoked exercise in the participants with a spinal cord injury. This finding supports the idea that a blood-borne compound and/or an elevated venous return is responsible for the exercise tachycardia, especially when neural feed-forward and feedback mechanisms are impaired (3).

In this study, leg musculature was separated from systemic circulation by inflation of thigh cuffs. For the able-bodied participants, this resulted in a reduction in oxygen uptake and carbon dioxide elimination. In both groups, there was an increase in these variables after release of the thigh cuffs. The increase was most pronounced in the participants with spinal cord injury, suggesting that a metabolic factor arising from the working limbs contributes to the exercise ventilation (1, 10, 11). Indeed, during electrically stimulated exercise, pH falls and potassium is elevated more than during comparable voluntary exercise (13), both of which can stimulate ventilation (25). The fact that the ventilatory response to exercise was prominent in the individuals with spinal cord injury provides no evidence for neural signals from the exercising limbs dominating ventilation during exercise. Recently, it has been suggested that pressure- and/or perfusion-related afferent signals within skeletal muscle contribute to regulation of respiration (9). Interestingly, in the present study, during postischemia after exercise, the two groups differed in their ventilatory response. This indicates that the effect of releasing the cuff and thus altering tissue pressure and flow did result in an acutely differential response in the two groups.

Vascular occlusion reduced the heart rate response to exercise in most of the patients with spinal cord injury and in able-bodied individuals and therefore provides further evidence for the importance of humoral feedback for the exercise tachycardia. This hypothesis is based on the fact that both neural feed-forward and feedback mechanisms are severely impaired or absent in the participants with spinal cord injury. That the cuff application in control individuals per se could contribute to the heart rate response cannot be totally excluded (Table 1). However, the above hypothesis is further supported by the finding that a reduction in heart rate took place during ischemic exercise in the able-bodied humans, although the discomfort evoked by such procedure would be assumed to increase heart rate. An early suggestion was that the heart rate response to exercise is determined by an elevated blood temperature (15). The suggestion has gained little support mainly because a change in temperature is inadequate to explain the almost instantaneous increase in heart rate at the onset of exercise (2). Evoked exercise with free circulation to the legs in spinal cord-injured individuals increases cardiac output to ~10 l/min and oxygen uptake to 1 l/min (7, 19), and the central blood volume would be expected to be elevated. Thus the absent heart rate response during ischemic exercise could reflect the effect of a reduced venous return, and, when Brown et al. (5) occluded venous return during stimulated exercise, blood pressure was unchanged in most of their participants, with a decrease in heart rate. With arterial occlusion, blood pressure increased markedly in two of the spinal cord-injured individuals, and we cannot exclude the possibility that their heart rate response is dominated by stimulation of volume- and baroreceptors. However, an elevation of the central blood volume by itself (during head-down tilt; the Trendelenburg position) does not affect heart rate (19).

To identify a possible metabolic factor, another experiment was performed involving the same spinal cord-injured individuals and the same exercise protocol. A twofold increase in plasma calcitonin gene-related peptide could be found compared with the value in healthy humans (unpublished data from Ref. 17), and this factor may contribute to the increase in heart rate (23). Also a three- to fivefold increase in plasma norepinephrine was detected, which reflected spillover from the working muscles. No increase in circulating catecholamines is therefore expected during exercise with an arterial cuff, but catecholamines would be expected to increase after release of the cuffs when heart rate also increased. Both in this phase and during exercise, the higher heart rate in the spinal-injured participants compared with the able-bodied subjects is consistent with the finding that spinal cord-injured individuals are hypersensitive to norepinephrine (16).

Electrically stimulated cycling induces an increase in plasma lactate that is higher than during voluntary cycling (12), but it is unlikely that lactate and/or the reduced pH could trigger a heart rate response (22). Of the other metabolites that are increased during exercise, potassium would be expected to decrease rather than to increase heart rate, and it would have the opposite effect on ventilation (20). Further evidence for a metabolic substance regulating heart rate in the participants with spinal cord injury is gained from the slower recovery in heart rate. In the able-bodied subjects, recovery heart rate was consistently below the lower 95% confidence limit of the exercise value after ~1 s, whereas it was statistically unchanged during the first 30 s in the para- and tetraplegic subjects.

Taken together, the increase in heart rate during electrically stimulated cycling in persons with spinal injury and the reduced or abolished heart rate response to volitional or evoked cycling with the circulation to the legs occluded both suggest that humoral feedback is of importance for the heart rate response to exercise. This is especially the case when influence from the central nervous system and neural feedback from the working muscles are impaired.