The major finding of the present study was that MSNA BF fell during leg cycling at mild intensity when the pedaling frequency was increased (thus enhancing muscle pump). This result supports the hypothesis that no change or a decrease in sympathetic vasomotor outflow during dynamic leg exercise up to mild intensity may be due to muscle pump-induced increase in central blood volume, thereby loading the cardiopulmonary baroreceptors. The results from this study provide additional information concerning the mechanism of the MSNA response to dynamic leg exercise performed at mild intensity.
Effect of the muscle pump-induced increase in central blood volume on MSNA
Several previous studies have attempted to evaluate changes in MSNA during dynamic leg exercise, and revealed that MSNA tended to decrease or was unchanged at light and mild exercise intensities compared with that at rest (Seals et al. 1988; Saito and Mano 1991; Ray et al. 1993; Saito et al. 1993; Callister et al. 1994; Ichinose et al. 2008; Katayama et al. 2011). Similarly, in the present study, MSNA BF during exercise in the 60EX trial showed little change from resting values, as shown in Fig. 4. Also, MSNA BI decreased during exercise in the 60EX trial (Table 1). One possible explanation for this observation is that dynamic leg exercise-induced muscle pump was accompanied by an increase in central blood volume. Ray et al. (1993) compared MSNA responses to dynamic a one-legged knee exercise performed in the sitting and supine positions. They compared MSNA response to dynamic exercise between supine and upright positions and noted that changes in central blood volume affected the MSNA response to dynamic exercise. In addition, Vollianitis and Secher (Volianitis and Secher 2002) reported that arterial BP was reduced to a level below that achievable upon arm exercise alone when leg-cycling was added to an arm-cranking exercise. These results suggest that leg muscle pump strongly affects BP rather than the central command or the exercise pressor reflex. In addition, changing central blood volume affected peripheral vascular resistance (Donald and Shepherd 1978), and the extent of CO is negatively correlated with MSNA (Charkoudian et al. 2005). From these findings, it would be expected that loading of cardiopulmonary baroreceptors by a muscle pump-induced increase in central blood volume would suppress sympathetic vasomotor activity during dynamic leg exercise (Ray et al. 1993; Ray and Saito 1999; Fadel and Raven 2012; Raven and Chapleau 2014).
In the present study, we utilized a traditional method (an increase in pedal frequency from 60 to 80 rpm) to further enhance skeletal muscle pump frequency (Gotshall et al. 1996; Rowland and Lisowski 2001; Ogoh et al. 2007). This manipulated central blood volume in a controlled manner; SV and CO increased significantly during leg cycling at 80 rpm (the 80EX trial) compared with the values recorded at 60 rpm (the 60EX trial) (Fig. 1). Since CO is equal to venous return under steady-state conditions (Badeer 1981; Young 2010), increased CO during the 80EX trial indicates enhanced central blood volume. Moreover, thoracic impedance, an index of central blood volume, was higher during the 80EX trial than the 60EX trial (Fig. 2). Consequently, MSNA BF decreased during the 80EX trial (Fig. 4), and MSNA BI (Table 1) during this trial was lower than during the 60EX trial, suggesting that enhanced muscle pump, which caused loading of cardiopulmonary baroreceptors, could inhibit sympathetic vasomotor outflow during dynamic leg exercise at mild intensity.
In an animal study, the BP response at the onset of exercise was not maintained when the cardiopulmonary receptors operated alone (without arterial baroreflex) (Walgenbach and Donald 1983). They concluded that cardiopulmonary receptors do not have a significant role in BP regulation. However, Ogoh et al. (2007) reported that loading of cardiopulmonary baroreceptors modified the arterial baroreflex evident during steady-state exercise. In addition, orthostatic stress increased peripheral vascular resistance via unloading of cardiopulmonary baroreceptors to maintain the required BP during mild-intensity exercise (Mack et al. 1988). These earlier findings indicated that the cardiopulmonary and arterial baroreflexes interacted to regulate BP during exercise. Similarly, in the present study, sympathetic vasoconstriction outflow was suppressed by loading of cardiopulmonary baroreceptors during the 80EX trial. Interestingly, BP during exercise in the 80EX trial did not differ from that in the 60EX trial (Table 1). Therefore, suppression of sympathetic vasoconstriction outflow by the cardiopulmonary reflex in the 80EX trial may prevent overshooting of BP during exercise. Taking these observations into consideration, cardiopulmonary receptors could play an important role for sympathetic vasomotor outflow to maintain adequate BP against changes in central blood volume during steady-state mild dynamic exercise.
We need to consider other possible mechanisms affecting MSNA during dynamic leg exercise. First, it is necessary to consider energy expenditure which is related to central command and the exercise pressor reflex. It is well known that an increase in pedaling cadence induces a rise in gross energy expenditure at a constant load because internal work is altered (Lollgen et al. 1980; Wells et al. 1986). Thus, increases in cycle pedaling cadence may produce greater activation of central command and the exercise pressor reflex when workload is kept constant. In order to apply the same central command and exercise pressor reflex, we used differential workloads to obtain the same energy expenditure (i.e., VO2) during the 60EX and 80EX trials (Table 1) (Ogoh et al. 2007). Similar to VO2, there were no significant differences in HR and RPE between the two trials. Therefore, we suppose that central command inputs were similar during two trials types; although, it is difficult to estimate the extent thereof.
In terms of the exercise pressor reflex, the metaboreflex and mechanoreflex within skeletal muscle contribute to changes in MSNA during exercise. In an animal study, it has been revealed that group IV muscle afferent was increased during low intensity dynamic exercise (Adreani et al. 1997). An increase in muscle contraction frequency may limit muscle blood flow, thereby triggering metabolite accumulation. However, Ferreira et al. (2006) found no difference in the extent of vastus lateralis deoxygenation (assessed using near–infrared spectroscopy) during cycling at 60 and 100 rpm, indicating that the increase in contraction frequency did not impair blood flow to the muscle. Further, the VCO2 value obtained during the 80EX trial did not differ from those of the 60EX trial (Table 1). Collectively, although metaboreflex during dynamic leg cycling at low intensity could enhance MSNA, it seems likely that the extent of metaboreflex did not differ between the 60EX and 80EX trials. As for the mechanoreflex, mechanoreceptor loading certainly increases as muscle contraction frequency rises, and group III muscle afferent was enhanced during dynamic exercise at low intensity (Adreani et al. 1997). Passive stretching of the hindlimb muscle significantly increased renal sympathetic nerve activity (Matsukawa et al. 1990). In addition, MSNA has been reported to increase during passive muscle stretching in humans (Cui et al. 2006). These findings suggest that an increase in mechanoreceptor loading during the 80EX trial could enhance, not inhibit, MSNA.
Finally, respiratory modulation of MSNA should be mentioned (Hagbarth and Vallbo 1968; Eckberg et al. 1985; Dempsey et al. 2002). In the present study, subjects adopted a high breathing frequency during the 80EX trial; however, this difference was small. Seals et al. (1990) showed that differences in depths and patterns of breathing influenced the within-breath MSNA modulation. However, St Croix et al. (2000) found that MSNA was unchanged when the breathing frequency increased threefold (via voluntary hyperpnea). Therefore, it is unlikely that the slight difference in breathing frequency between the 60EX and 80EX trials significantly affected MSNA; although, we cannot entirely role out respiratory modulation of MSNA.
Technical considerations and limitations
Several technical considerations and limitations of the study should be noted. One limitation was the use of thoracic electrical impedance as an index of central blood volume; we could not obtain direct measurements of central venous pressure. Previous studies found that changes in thoracic impedance relates to central venous pressure during lower body positive or negative pressure (Ebert et al. 1986; Cai et al. 2000). As in the present study, Ogoh et al. (2007) found that thoracic impedance decreased when the pedal cadence rose from 60 to 80 rpm (at the same VO2) during submaximal leg cycling at mild intensity. In addition to thoracic impedance, we measured the central hemodynamic variables SV and CO by both impedance cardiography and a Doppler echocardiographic technique. As CO is equal to the venous return under steady-state conditions (Badeer 1981; Young 2010), we presumed that changes in central blood volume were reflected in the CO values. Although the absolute values of SV and CO obtained by the Doppler technique were lower than those derived by the impedance method (in line with data of a previous study) (Christie et al. 1987), both SV and CO were significantly higher during the 80EX trial (Fig. 1 and Table 2). Taking these observations into consideration, it seems reasonable to suggest that central blood volume increased during the 80EX trial of the present study.
It is necessary to consider exercise intensity and duration. We utilized mild exercise intensity (40% VO2peak at 60 rpm), because the percentage of successful MSNA recordings is high when the exercise intensity is mild and movements of the arm and the body become greater during leg cycling with a high pedaling rate (Katayama et al. 2012, 2013). The extent of contribution of cardiopulmonary baroreceptor to sympathetic vasoconstriction outflow would alter with increasing exercise intensity, as well as other mechanisms (e.g., metaboreflex) (Ray et al. 1993; Fisher et al. 2005; Boushel 2010). As for the exercise duration, Saito et al. (1997) and Ray et al. (1993) demonstrated a progressive increase in MSNA during 30- and 40-min of dynamic leg exercise at mild intensity. Therefore, it may be that the influence of the cardiopulmonary receptors overwhelms when exercise intensity is high and exercise duration prolonged (Ray et al. 1993).
Another limitation of this study was the limited number and characteristics of subjects. Clinically, cardiopulmonary baroreceptor is particularly important to regulate arterial BP at rest and during exercise. It has been reported that resting MSNA in obesity and patients with heart failure and hypertension is higher than in healthy individuals (Yamada et al. 1989; Grassi et al. 2003; Witte et al. 2008). Furthermore, cardiopulmonary reflex is impaired in patients with heart failure (Dibner-Dunlap et al. 1996). Thus, it is assumed that the MSNA response to changes in central hemodynamics during dynamic leg exercise in patients differs from that in healthy subjects.