On the contribution of group III and IV muscle afferents to the circulatory response to rhythmic exercise in humans

Non-technical summary

We investigated the role of thin fibre muscle afferents in the circulatory response to exercise in humans. The results not only document the importance of continuous afferent feedback from working human skeletal muscle to achieve appropriate haemodynamic and ventilatory responses to exercise but also suggest that the relative contribution of this mechanism is larger than traditionally accepted.

Abstract

Abstract

We investigated the role of skeletal muscle afferent feedback in circulatory control during rhythmic exercise in humans. Nine healthy males performed single leg knee-extensor exercise (15/30/45 watts, 3 min each) under both control conditions (Ctrl) and with lumbar intrathecal fentanyl impairing μ-opioid receptor-sensitive muscle afferents. Cardiac output and femoral blood flow were determined, and femoral arterial/venous blood samples were collected during the final minute of each workload. To rule out cephalad migration of fentanyl to the brainstem, we documented unchanged resting ventilatory responses to different levels of hypercapnia. There were no haemodynamic differences between conditions at rest. However, during exercise cardiac output was ∼20% lower with fentanyl blockade compared to control (P < 0.05), secondary to a 6% and 13% reduction in heart rate and stroke volume, respectively. Throughout exercise mean arterial pressure (MAP) was reduced by 7% (P < 0.01) which is likely to have contributed to the 15% fall in femoral blood flow. However, MAP was not completely responsible for this peripheral haemodynamic change as vascular conductance was also attenuated (∼9%). Evidence of increasing noradrenaline spillover (P = 0.09) implicated an elevation in sympathetic outflow in this response. The attenuated femoral blood flow during exercise with fentanyl was associated with a 17% reduction in leg O₂ delivery (P < 0.01) and a concomitant rise in the arteriovenous O₂ difference (4–9%), but leg O₂ consumption remained 7–13% lower than control (P < 0.05). Our findings reveal an essential contribution of continuous muscle afferent feedback to ensure the appropriate haemodynamic and ultimately metabolic response to rhythmic exercise in humans.

Introduction

Autonomic and ventilatory responses to exercise are primarily regulated by two, largely separate systems. The first, a feedforward mechanism termed ‘central command’, elicits cardiovascular and ventilatory responses to exercise (Waldrop et al. 1996) presumably via the activation of insular cortex and/or medial prefrontal cortex (Williamson et al. 2006). The second, a feedback mechanism, reflexly changes ventilation and autonomic responses as a consequence of limb muscle contraction (Kaufman & Forster, 1996). The sensory arm of this reflex arc consists of thin fibre muscle afferents (group III and IV) which are sensitive to both mechanical and metabolic stimuli within the exercising muscle (Kaufman et al. 1983). However, despite numerous investigations throughout the 20th century, the relative importance and contribution of these regulatory mechanisms to the cardiovascular and ventilatory response during exercise remains unclear.

We have provided evidence, in the intact human, of a substantial role for thin fibre muscle afferents in regulating ventilatory, heart rate (HR), and pressor responses during constant-load bicycle exercise including mild and heavy intensities (Amann et al. 2010). A key finding from this prior work was that continuous neural feedback is required from rhythmically working limb muscles to the spinal cord and/or brainstem (perhaps a tonic input) in order to ensure adequate ventilatory and cardiovascular responses throughout exercise. This contrasts with the view which depicts afferent feedback as a temporary error signal (O'Leary & Sheriff, 1995) to the brainstem reporting an acute mismatch between blood/oxygen supply and demand (Mitchell et al. 1983).

A continuous neural feedback from working limb muscles to maintain an appropriate HR and pressor response to exercise (Amann et al. 2010) supports the concept that there is likely to be a role for muscle afferents in the control of skeletal muscle blood flow (Alam & Smirk, 1937; O'Leary & Sheriff, 1995). Specifically, it has been argued that with rhythmic exercise, during which cardiac output () can be increased, the effects of muscle afferents on the CNS might evoke an increase in blood flow to the working muscles by increasing perfusion pressure secondary to an increase in and resistance in inactive vascular beds (Boushel, 2010). In contrast, during rhythmic exercise, during which cannot be increased to match increases in vascular conductance in working muscle, the reflex effects of muscle afferents might limit skeletal muscle blood flow by evoking sympathetically mediated vasoconstriction (Boushel, 2010).

Although findings from recent studies on chronically instrumented dogs suggest an involvement of afferent feedback from working limb muscle in raising and, importantly, maintaining adequate skeletal muscle blood flow during treadmill running (O'Leary & Sheriff, 1995), evidence of this concept is lacking in humans. Consequently, we conducted a set of experiments to investigate the role of group III/IV muscle afferents, with their receptive fields in the lower limb, in determining skeletal muscle blood flow during exercise in healthy humans. We utilized the single leg knee-extensor model, an exercise modality in which exercising quadriceps blood flow (i.e. vascular conductance) cannot outstrip (Andersen & Saltin, 1985) and therefore only evokes minimal sympathetically mediated vasoconstriction in the working limb (Saltin et al. 1998). Lumbar intrathecal fentanyl was administered to partially block lower limb muscle afferents and we tested the hypotheses that limiting muscle afferent feedback during knee-extensor exercise would (1) attenuate central haemodynamic responses (, HR, and stroke volume (SV)), (2) attenuate the exercise pressor reflex, and subsequently (3) attenuate femoral blood flow (FBF).

Methods

Nine recreationally active males volunteered to participate in the study (age 27 ± 4 years, body weight 77 ± 13 kg, height 1.77 ± 0.07 m, right thigh muscle mass 2.6 ± 0.3 kg). Written informed consent was obtained from each participant. All procedures conformed to the Declaration of Helsinki and were approved by the Institutional Review Board, University of Utah.

Experimental protocol

All participants were thoroughly familiarized with all experimental procedures. Between 48 and 72 h following their final practice session, subjects returned to the laboratory where their right femoral artery and vein were catheterized (18 gauge central line catheters, Arrow International, Reading, PA, USA) using the Seldinger technique. Following a 30 min rest period, CO₂ sensitivity was evaluated by determining the ventilatory response to three levels of inspiratory CO₂ () while comfortably sitting on the knee-extensor ergometer. Following a short break, control (Ctrl) values for FBF and pulmonary/cardiovascular variables were obtained and resting arterial/venous blood samples were taken. The Ctrl trial consisted of three levels of constant load knee-extensor exercise with the right leg (15/30/45 W, 3 min each, 60 rev min⁻¹), each of which equated to less than 60% of the subjects’ maximal workload. Pulmonary variables, and FBF were recorded continuously and arterial/venous blood samples were taken during the final minute of each workload. Following a 2 h rest period, the subjects were placed in an upright seated position and intrathecal fentanyl (0.025 mg ml⁻¹), an opioid analgesic with no effect on the force generating capacity of the quadriceps (Amann et al. 2009, 2010), was delivered at vertebral interspace L3–L4 (Amann et al. 2009). To minimize the potential risk of cephalad movement within the cerebrospinal fluid (CSF), subjects remained in the upright seated position throughout the remainder of the study. Cutaneous hypoaesthesia to pinprick and cold perception on the torso and upper limbs were examined prior to exercise. Approximately 20 min post-fentanyl injection, the steady-state CO₂ response test, baseline measures, and identical exercise were repeated. In all subjects the time duration from the fentanyl injection to the end of the experiment was less than 60 min.

Measurements

FBF

Simultaneous measurements of common femoral arterial blood velocity (V_mean) and vessel diameter were performed distal to the inguinal ligament and proximal to the bifurcation of the deep and superficial femoral arteries with a Logic 7 ultrasound system (General Electric Medical Systems, Fairfield, CT, USA). Using arterial diameter and V_mean, FBF was calculated as: FBF = V_meanπ(vessel diameter/2)²× 60.

Pulmonary and cardiovascular responses

Ventilation and pulmonary gas exchange were measured continuously using an open circuit system (True Max 2400, Parvo Medics, Sandy, UT, USA). HR was measured from the R-R interval using a three-lead electrocardiogram (Biopac systems Inc., Goleta, CA, USA). SV was calculated from beat-by-beat pressure waveforms assessed by photoplethysmography (Beatscope version 1.1; Finapress Medical Systems, Amsterdam, The Netherlands) and was calculated as the product of SV and HR. Arterial and venous blood pressure measurements were collected continuously from within the femoral artery and vein, with pressure transducers (Transpac IV, Hospira, Lake Forest, IL, USA) placed at the level of the catheters. Mean arterial pressure (MAP) was calculated as diastolic pressure +⅓ (systolic pressure – diastolic pressure), mean femoral venous pressure (MVP) was the average of systolic and diastolic pressure. Leg vascular conductance (LVC) was calculated as FBF/(MAP – MVP).

Blood derived variables

Femoral arterial and venous blood samples were anaerobically collected and analysed (GEM 4000; Instrumentation Laboratory Co., Bedford, MA, USA). Blood gases were not corrected for blood temperature. Arterial () and venous () blood O₂ content were 1.39 (Hb) × (oxyhaemoglobin saturation/100) + 0.003 ×. Percentage O₂ extraction was calculated as: [(–)/]× 100. Oxygen delivery was calculated as the product of FBF and , and muscle as the product of – difference and FBF. Plasma noradrenaline (NA) and adrenaline concentrations were measured in duplicate by a competitive ELISA (coefficient of variation: 19%; 2-CAT ELISA; Labor Diagnostika Nord GmbH & Co. KG, Nordhorn, Germany). The rate of NA spillover into plasma was determined as:

where C_v and C_a are femoral venous and arterial plasma NA concentrations, Epi_e is the fractional extraction of adrenaline, and LPF is leg plasma flow determined from FBF and haematocrit (Savard et al. 1989).

Steady-state CO₂ response test

Measurements were carried out using a steady-state, open circuit technique (Berkenbosch et al. 1989). In addition to eupnoeic air breathing (5 min), ventilatory responses to two different concentrations of CO₂ (70% O₂, 3 and 6% CO₂, balance N₂) were measured in all subjects. The subjects breathed each gas mixture for 4 min and the tests were separated by at least 5 min of exposure to room air to allow ventilatory variables to return to baseline levels. Arterial blood gases were collected during the final 30 s of each condition and analysed for . Breathing frequency (f_R) and tidal volume (V_T) were assessed and averaged over the final minute of each condition.

Statistical analysis

A two-way analysis of variance with repeated measures was performed to evaluate differences between trials. A least significance difference test identified the means that were significantly different with P < 0.05. Results are expressed as means ± SEM.

Results

Resting ventilatory responses to CO₂ and cutaneous hypoaesthesia

Table 1 contains the respiratory variables measured during the last minute of exposure to three different levels of under Ctrl and 15–20 min after fentanyl injection. Eupnoeic air breathing was not altered from Ctrl by fentanyl, as indicated by the nearly identical breathing patterns and very similar values in all nine subjects. Exposure to the two levels of increased resulted in similar values and subsequently hypercapnic ventilatory responses in both the Ctrl and fentanyl trial.

Table 1.

Ventilatory response to CO₂ at rest

		Arterial (mmHg)		fR (breaths min−1)		VT (l)
(%)	(mmHg)	Ctrl	Fentanyl	Ctrl	Fentanyl	Ctrl	Fentanyl
Room	0.2 ± 0.0	36.0 ± 1.7	35.7 ± 1.8	11.8 ± 1.3	13.1 ± 1.1	1.0 ± 0.3	0.9 ± 0.2
3	19.2 ± 0.0	39.4 ± 1.6	39.0 ± 1.6	12.0 ± 1.0	13.3 ± 1.2	1.2 ± 0.2	1.1 ± 0.2
6	38.5 ± 0.1	45.0 ± 1.1	44.3 ± 1.1	15.7 ± 1.1	15.5 ± 1.4	1.4 ± 0.1	1.5 ± 0.1

All experiments were performed at a barometric pressure of 641 ± 3 mmHg. n = 9.

Neurological examinations just prior to the start of the exercise revealed cutaneous hypoaesthesia to pinprick and cold perception between T7 and T9. This was evident by sensory changes on the torso at, or below, T7 and by the absence of sensory changes on the upper limbs (demarcating T1 and above).

Central haemodynamic responses

At rest, HR, SV, and MAP were not different under Ctrl and fentanyl conditions (all P > 0.2; Figs 1 and 2). Within both condition, HR and increased with each incremental workload. However, compared to Ctrl, and HR were consistently 19–22% and 5–7% lower, respectively, during exercise with fentanyl and this difference was significant at each workload (Fig. 1). Under Ctrl, SV rose from rest to exercise, but remained unchanged with further increases in workload. In contrast, SV did not change from rest to exercise in the fentanyl trial. During exercise, fentanyl had an overall main effect on SV resulting in an 11–15% decrease compared to Ctrl. MAP was consistently 6–7% lower during exercise with fentanyl and this reduction was evident at all three workloads (all P < 0.001; Fig. 2B). MVP increased similarly from rest to exercise in both conditions and fentanyl blockade continued to have no effect on MVP across each workload (all P > 0.3; Fig. 2C). Figure 3 illustrates the consistency of individual HR, and MAP responses in the Ctrl trial and with fentanyl at rest and across workloads.

Figure 1. Cardiac output, heart rate and stroke volume at rest and during the final minute of knee-extensor exercise at each submaximal work rate.

The P value indicates the overall main effect of fentanyl. There was no interaction effect. *P < 0.05 vs. Control. †P < 0.05 vs. 15 W.

Figure 2. Femoral blood flow, leg mean arterial and venous pressure, and leg vascular conductance at rest and during the final minute of knee-extensor exercise at each submaximal work rate.

The P-value indicates the overall main effect of fentanyl. There was no interaction effect. *P < 0.05 vs. Control. #P = 0.06. †P < 0.05 vs. 15 W.

Figure 3.

Identity plots to illustrate individual subject responses at rest and during knee-extensor exercise at each submaximal work rate

Peripheral haemodynamic responses

At rest, FBF, NA spillover, lactate and LVC were similar in both Ctrl and fentanyl condition (all P > 0.4; Fig. 2). Throughout exercise, fentanyl had an overall main effect on FBF and LVC. The difference in FBF failed to reach significance at 15 W (P = 0.06), but fentanyl caused a substantial reduction (14–16%) in FBF at the two highest workloads (Fig. 2A). The individual consistency of the FBF responses are illustrated in Fig. 3A. Fentanyl had an overall main effect on LVC which was on average 8–10% lower during the exercise with fentanyl; however, the difference failed to reach significance at each individual workload (all P > 0.15; Fig. 2D). NA spillover increased from rest to exercise in both conditions. NA spillover was invariant across the three levels of submaximal knee-extensor exercise during Ctrl (P = 0.84) and during exercise with fentanyl blockade (P = 0.15). However, fentanyl blockade during exercise revealed a tendency towards a main effect (P = 0.09; Table 2). Arterial and venous lactates were similar in both Ctrl and the fentanyl block conditions; however, lactate efflux was lower during exercise with fentanyl at 15 and 30 W (Table 2).

Table 2.

Femoral arterial/venous O₂ transport, noradrenaline spillover, and gas exchange variables obtained at rest and during the final minute of exercise

	REST		15 W		30 W		45 W		ANOVA
	Ctrl	Fentanyl	Ctrl	Fentanyl	Ctrl	Fentanyl	Ctrl	Fentanyl	P value
Hb (g dl−1)	15.7 ± 0.2	15.5 ± 0.3	15.8 ± 0.3	15.8 ± 0.2	15.9 ± 0.3	15.9 ± 0.3	16.1 ± 0.2	15.9 ± 0.3	0.312
(ml dl−1)	21.3 ± 0.3	21.1 ± 0.4	21.5 ± 0.4	21.2 ± 0.4	21.6 ± 0.4	21.2 ± 0.4	22.0 ± 0.3	21.5 ± 0.4	0.010
(ml dl−1)	14.6 ± 0.9	14.9 ± 0.6	8.2 ± 0.3	6.7 ± 0.4**	7.2 ± 0.3	6.0 ± 0.4**	6.2 ± 0.3	5.0 ± 0.4**	<0.001
(%)	96.5 ± 0.3	96.6 ± 0.3	96.5 ± 0.3	96.1 ± 0.4	96.4 ± 0.3	95.9 ± 0.3	96.5 ± 0.2	96.2 ± 0.2**	0.035
(%)	67.4 ± 4.0	68.8 ± 2.4	36.9 ± 1.1	30.4 ± 1.6**	31.9 ± 1.3	27.0 ± 1.3**	27.1 ± 1.3	22.3 ± 1.3**	<0.001
(mmHg)	84.7 ± 2.5	87.6 ± 3.6	88.6 ± 3.2	84.8 ± 3.9	87.0 ± 1.7	83.2 ± 2.3*	89.9 ± 2.2	86.7 ± 1.5*	0.001
(mmHg)	35.6 ± 1.6	34.5 ± 1.9	34.7 ± 1.9	36.0 ± 1.9*	35.2 ± 1.4	36.0 ± 1.4*	34.1 ± 1.6	35.6 ± 1.6**	0.003
(mmHg)	40.7 ± 3.2	41.1 ± 2.0	25.0 ± 0.9	22.3 ± 0.9**	23.2 ± 0.7	21.2 ± 0.9*	21.3 ± 0.9	19.8 ± 1.1*	0.003
(mmHg)	42.7 ± 1.7	42.7 ± 1.9	55.3 ± 2.4	54.0 ± 3.0	58.0 ± 2.1	59.3 ± 2.3	65.1 ± 2.5	69.6 ± 3.0	0.418
O2 extraction (%)	30.4 ± 2.9	29.2 ± 2.2	61.9 ± 1.0	68.5 ± 1.5**	66.9 ± 1.3	71.9 ± 1.3**	72.0 ± 1.3	76.8 ± 1.3**	<0.001
NA spillover (ng min−1)	94 ± 11	91 ± 21	340 ± 126	393 ± 177	334 ± 179	512 ± 237	339 ± 182	603 ± 234	0.092
Arterial pH	7.43 ± 0.01	7.43 ± 0.01	7.42 ± 0.02	7.43 ± 0.03	7.41 ± 0.01	7.39 ± 0.02	7.40 ± 0.01	7.38 ± 0.01**	0.262
Venous pH	7.39 ± 0.01	7.39 ± 0.02	7.31 ± 0.02	7.32 ± 0.02	7.29 ± 0.01	7.28 ± 0.01	7.25 ± 0.02	7.22 ± 0.01*	0.323
Arterial lactate (mmol l−1)	0.6 ± 0.0	0.6 ± 0.1	1.3 ± 0.2	1.3 ± 0.2	1.8 ± 0.3	1.8 ± 0.4	2.7 ± 0.5	2.9 ± 0.5	0.514
Venous lactate (mmol l−1)	0.7 ± 0.1	0.8 ± 0.1	2.1 ± 0.4	1.9 ± 0.4	2.5 ± 0.6	2.5 ± 0.6	4.2 ± 0.9	4.4 ± 0.9	0.961
Lactate efflux (mmol min−1)	0.1 ± 0.0	0.1 ± 0.0	2.3 ± 0.8	1.6 ± 0.6*	2.7 ± 0.8	2.2 ± 0.8*	6.0 ± 1.5	5.4 ± 1.3	0.067

All experiments were performed at a barometric pressure of 641 ± 4 mmHg.

^*P < 0.05

^**P < 0.01vs. Ctrl.

Leg O₂ supply and O₂ utilization

At rest, O₂ delivery, arteriovenous O₂ difference and leg were similar in both conditions (all P > 0.4). Throughout the exercise with fentanyl, O₂ delivery was lower at each workload (Fig. 4A). Arteriovenous O₂ difference was consistently higher (4–9%) during exercise with fentanyl (Fig. 4B) due to a substantial reduction in (Table 2). However, the compensatory rise in arteriovenous O₂ difference (secondary to the fall in FBF) failed to restore leg to that of Ctrl conditions, revealing an overall main effect of afferent blockade on leg . However, this reduction in leg failed to reach significance at each individual work rate (15 W, P = 0.42; 30 W, P = 0.12; and 45 W, P = 0.07) (Fig. 4C).

Figure 4. Leg O₂ supply and demand at rest and during the final minute of knee-extensor exercise at each submaximal work rate.

The P value indicates the overall main effect of fentanyl. There was no interaction effect. *P < 0.05 vs. Control. †P < 0.05 vs. 15 W.

Ventilatory responses

At rest, fentanyl had no effect on , breathing pattern, arterial blood gases, or haemoglobin saturation (Table 3). Throughout the exercise, fentanyl had an overall main effect on ; however, the reduction in ventilation during exercise with fentanyl failed to reach significance at each individual work rate (15 W P = 0.08), 30 W (P = 0.06), and 45 W (P = 0.09; Table 3). This consistent hypoventilation caused a small but significant increase in at all workloads, and a reduction in and arterial haemoglobin saturation throughout exercise (Table 2).

Table 3.

Ventilatory responses at rest and during the final minute of single knee-extensor exercise

	REST		15 W		30 W		45 W		ANOVA
	Ctrl	Fentanyl	Ctrl	Fentanyl	Ctrl	Fentanyl	Ctrl	Fentanyl	P value
(l min−1)	9.3 ± 0.7	9.8 ± 1.4	22.8 ± 2.4	21.9 ± 2.6	26.9 ± 1.2	25.4 ± 1.8	37.5 ± 3.2	34.4 ± 2.2	0.016
fR (breaths min−1)	12.7 ± 1.4	13.8 ± 1.3	21.3 ± 1.6	21.4 ± 1.7	24.6 ± 1.4	22.9 ± 1.9	28.2 ± 1.9	26.5 ± 2.1	0.106
VT (l)	0.80 ± 0.19	0.80 ± 0.21	1.13 ± 0.20	1.12 ± 0.21	1.14 ± 0.09	1.15 ± 0.13	1.35 ± 0.09	1.37 ± 0.14	0.352
(l min−1)	0.27 ± 0.02	0.28 ± 0.02	0.76 ± 0.03	0.77 ± 0.05	0.95 ± 0.03	0.93 ± 0.04	1.19 ± 0.05	1.16 ± 0.05	0.566
(l min−1)	0.22 ± 0.02	0.23 ± 0.03	0.64 ± 0.04	0.59 ± 0.05	0.82 ± 0.03	0.77 ± 0.04	1.13 ± 0.05	1.07 ± 0.05	0.029
RER	0.81 ± 0.02	0.83 ± 0.04	0.83 ± 0.03	0.77 ± 0.01*	0.86 ± 0.01	0.83 ± 0.02*	0.95 ± 0.02	0.92 ± 0.01*	0.002
/	34.1 ± 1.1	35.2 ± 2.1	29.1 ± 2.2	28.3 ± 2.2	28.1 ± 0.8	27.6 ± 1.4	31.3 ± 2.1	29.6 ± 1.4	0.059
/	41.4 ± 1.2	41.8 ± 0.9	34.7 ± 1.7	36.8 ± 1.9	32.9 ± 1.1	33.5 ± 1.8	32.8 ± 1.6	32.2 ± 1.3	0.178

The P-value indicates the overall main effect of fentanyl during exercise.

^*P < 0.05 vs Ctrl.

Discussion

This study sought to evaluate the role of muscle afferent feedback in the circulatory control during rhythmic exercise in humans. We used lumbar intrathecal fentanyl to block the central projection of μ-opioid-receptor sensitive group III/IV muscle afferents from the lower limbs during single leg knee-extensor exercise ranging from mild to moderate intensities. At rest, the afferent blockade had no effect on central and peripheral haemodynamics or ventilation. However, throughout exercise with fentanyl, was attenuated, secondary to both a lower HR and SV, and this was accompanied by a substantial reduction in MAP. Not only was FBF attenuated during the exercise with fentanyl, but LVC was also reduced compared to Ctrl. The attenuated FBF during exercise with fentanyl resulted in a reduction in leg O₂ delivery and a concomitant rise in arteriovenous O₂ difference, but leg remained lower than during Ctrl exercise. Taken together, these findings provide evidence, in humans, of a critical role of group III/IV muscle afferents in the regulation of the circulatory and ultimately metabolic response to rhythmic exercise.

Experimental use of intrathecal fentanyl

We used a lumbar intrathecal bolus injection of the selective μ-opioid receptor agonist fentanyl to attenuate the central projection of lower limb muscle afferents synapsing on cells in laminae I and V of the lumbar dorsal horn of the spinal cord (Wilson & Hand, 1997). These dorsal horn cells project to the ventral lateral medulla and the nucleus tractus solitarii where they influence breathing and cardiovascular control (Craig, 1995). Stimulation of spinal opioid receptors inhibits group III/IV mediated input to the spinal cord (Yaksh & Noueihed, 1985; Meintjes et al. 1995; Kalliomaki et al. 1998) without affecting the force generating capacity of skeletal muscle (Amann et al. 2009, 2010). Intrathecally applied opioid receptor agonists have been documented to reduce HR, MAP and ventilatory responses to leg cycling exercise in healthy humans (Amann et al. 2010), to isometric limb contractions in anaesthetized cats (Hill & Kaufman, 1990; Meintjes et al. 1995), and to ischaemic exercise in awake dogs (Pomeroy et al. 1986).

Previous human (Eisenach et al. 2003) and animal (Swenson et al. 2001) studies reveal the potential for cephalad movement of opioids within the CSF. Individual variations in the extent of the ‘upstream’ spread have been reported, but it is thought that axial spinal length (distance between injection site and brainstem) is likely to play a key role (Swenson et al. 2001). Furthermore, as fentanyl is slightly hypobaric relative to CSF, patient position will also contribute to the cephalad movement (Swenson et al. 2001). A migration of fentanyl sufficient to reach the brainstem would certainly negate the significance of our findings because the direct effect of fentanyl on medullary opioid receptors is recognized to affect neurons involved in cardiovascular (Sun et al. 1996; Caringi et al. 1998) and ventilatory (Lalley, 2008) control. We had to exclude two of the seven subjects in our previous study due to suspicion of a direct effect of fentanyl on medullary opioid receptors (Amann et al. 2010). Aware of these important issues, we reduced the concentration of fentanyl in comparison to our previous study (Amann et al. 2010) and subjects were kept upright at all times and again individually assessed via neurological examination and ventilatory responsiveness to hypercapnia ∼20 min following fentanyl delivery. The negative outcome of this battery of tests provides evidence for a segmental effect of the drug and excludes any direct medullary impact of fentanyl in each of our subjects.

Central haemodynamics and group III/IV muscle afferents

Thin fibre muscle afferents evoke inotropic and chronotropic effects on the heart to increase in exercising animals (O'Leary & Augustyniak, 1998; Sheriff et al. 1998) and humans (Saltin et al. 1998; Shoemaker et al. 2007; Boushel, 2010). Specifically, these sensory neurons appear to reflexly increase HR (O'Leary & Augustyniak, 1998), and maintain or increase right atrial pressure (Sheriff et al. 1998) and ventricular contractility (Mitchell et al. 1977; O'Leary & Augustyniak, 1998) during exercise. The adjustments in myocardial contractility combine to maintain SV in the face of tachycardia and associated decreases in cardiac filling time. These reflex-dependent modifications allow and subsequently MAP to rise during submaximal exercise which depicts a key determinant of a pressor response and increases in blood flow to the working muscle. The current observation of a reduced , secondary to reduced HR and SV, during exercise with fentanyl confirms these functions of group III/IV muscle afferents. Interestingly, SV increased at the onset of Ctrl exercise and remained at this elevated level throughout exercise (Fig. 1). In contrast, SV was unchanged from resting levels during exercise with fentanyl. It should be noted that potential differences in venous return as the cause of the lower SV during exercise with fentanyl compared to Ctrl cannot be excluded at this point. However, the fact that SV did not fall below resting levels, due to the exercise-induced tachycardia and the reduction in ventricular filling time during the fentanyl trial, reveals some remaining capacity to increase myocardial contractility during exercise despite attenuated group III/IV muscle afferent feedback. The rather incomplete afferent blockade associated with the use of fentanyl (limited to μ-opioid receptor-sensitive afferents) and the ∼20% increase in femoral venous pressure during exercise might contribute to this observation.

As already discussed, it is likely that the missing effect of the muscle afferents on accounts for the reduced MAP during the exercise with fentanyl. This interpretation is in agreement with previous human studies (Friedman et al. 1992; Shoemaker et al. 2007) and is supported by data from animal experiments revealing that, during submaximal exercise, the effects of muscle afferents on MAP are nearly exclusively mediated by changes in (Sheriff et al. 1998; Augustyniak et al. 2001). Thus, our observations support the idea that the reflex-induced effects on MAP are predominantly mediated via , but contrast with another study claiming that MAP is mainly mediated by the effects of muscle afferents on vascular resistance (Ray et al. 1994).

Although the current data are in agreement with previous human and animal studies documenting a significant effect of thin fibre muscle afferents on the heart, it is important to emphasize that we did not directly measure SV and . Instead, we estimated SV using the Modelflow approach (Wesseling et al. 1993; based on the arterial pressure waveform from the Finapress), which has recently been found to potentially underestimate when there is a substantial reduction in MAP (Shibasaki et al. 2011). Therefore, the agreement of our data with previous findings might be, in part, due to this limitation of the Modelflow approach.

Peripheral haemodynamics and group III/IV muscle afferents

Our experiments reveal a substantial contribution of muscle afferent feedback on limb blood flow in exercising human (Fig. 3). Previous human research designed to address the effects of muscle afferents on the regulation of skeletal muscle blood flow has focused on facilitating afferent feedback by occluding blood flow to the exercising muscle (Joyner, 1991; Rowell et al. 1991). These investigations have relied on indirect measures of blood flow, namely oxygen saturation of the venous drainage. This method is limited (O'Leary & Sheriff, 1995), which may have contributed to the contradictory conclusions regarding the role of muscle afferents in blood flow regulation. Other human studies have focused on blocking the central effects of muscle afferents during both electrically evoked (Strange et al. 1993) and voluntary (Kjaer et al. 1999) leg exercise using a local anaesthetic (lumbar epidural bupivacaine). The results of these experiments have also been equivocal, documenting a reduced, or unchanged leg blood flow response to exercise with complete (Strange et al. 1993), or partial (Kjaer et al. 1999), spinal neurotransmission blockade. Furthermore, two major issues prevent a clear interpretation of each of these studies. First, during voluntary exercise (Kjaer et al. 1999), local anaesthetics attenuate efferent as well as afferent nerve activity and such a drug-induced ‘muscle weakening’ inevitably requires an increase in central motor drive (i.e. feedforward) in order to maintain a given external workload. Thus, even a partial spinal block evoked by a local anaesthetic creates a condition of reduced feedback in the face of increased feedforward (Kjaer et al. 1999) and this increase in central command, per se, augments the cardiovascular response to exercise (Galbo et al. 1987; Williamson et al. 1996; Amann et al. 2008) blurring any potential conclusions regarding the role of afferent feedback. Second, electrical stimulation used to evoke muscle contractions not only causes an atypical motor unit recruitment, but also the electrical current may have directly stimulated afferent neurons confounding the results.

Therefore, to circumvent these experimental concerns, we employed voluntary exercise and used a μ-opioid receptor agonist which blocks group III/IV muscle afferents (Pomeroy et al. 1986; Hill & Kaufman, 1990; Meintjes et al. 1995; Amann et al. 2010) without affecting the force-generating capacity of skeletal muscle and the magnitude of central motor drive during rhythmic constant-load exercise (Amann et al. 2010). With this approach we have demonstrated that attenuated skeletal muscle afferent feedback results in a reduction in FBF in an exercising human limb (Fig. 3).

Of note, subsequent to the failure to appropriately increase FBF in the absence of the central effects of group III/IV muscle afferents during exercise, arteriovenous O₂ difference increased, but leg was not restored to the level recorded during Ctrl. Furthermore, arterial and venous lactate concentrations were similar in both conditions supporting earlier findings (Mortensen et al. 2009). Although these observations contrast with traditional thinking, namely that leg for a given work rate remains the same, there is a growing body of literature suggesting that, when faced with limited O₂ supply, skeletal muscle may, within certain limits, be able to titrate metabolic efficiency such that metabolic demand equals O₂ supply (Harms et al. 1997; Calbet et al. 2003; Mortensen et al. 2007, 2009; Olson et al. 2010; Larsen et al. 2011). Evidence in support of this being a global metabolic reduction in energy cost (Larsen et al. 2011) and not a shift toward a less oxidative pathway is supported by the trend for a reduction, and not an increase (as found by others; Mortensen et al. 2009), in lactate efflux across the leg and reduced RER (and ) in the blocked state (Tables 2 and 3). Alternatively, reduced fast twitch muscle fibre recruitment could have contributed to the lower leg at a given work rate with the fentanyl blockade (Krustrup et al. 2008). However, if lumbar intrathecal fentanyl did indeed result in the preferential recruitment of type I muscle fibres during exercise, this would also likely result in a reduced force generating capacity of the quadriceps muscle – but this was probably not the case (Grant et al. 1996; Standl et al. 2001; Amann et al. 2009, 2010). Despite being an unanticipated finding of the current study, these observations are germane and emphasize that alterations in skeletal muscle blood flow due to skeletal muscle afferent feedback are likely to have numerous consequences during exercise (e.g. Amann & Calbet, 2008).

A component of the reduction in FBF during fentanyl exercise can be explained by the attenuated rise in perfusion pressure evoked by the afferent blockade. These afferents are likely to play a role in regulating MAP during exercise by influencing (Friedman et al. 1992; Shoemaker et al. 2007), which in turn can affect FBF. However, what is intriguing from the current data is that with reduced afferent feedback both and MAP were consistently reduced, the latter of which would be expected to generate an error signal via the arterial baroreflex causing a subsequent increase in muscle sympathetic activity and, to a smaller degree, (Ogoh et al. 2003). Although somewhat circumstantial, there is evidence from this study suggesting that the baroreflex may have been attempting to restore MAP in this fashion during exercise with fentanyl, as indicated by the elevated NA spillover during the fentanyl trial and the attenuated LVC in comparison to Ctrl (Table 2 and Fig. 2). The rather incomplete attempt to restore MAP may be a consequence of the blocked afferent feedback which, per se, contributes to the resetting of the carotid baroreflex to operate at a higher arterial pressure (Smith et al. 2003). However, regardless of the mechanism, it is likely that these baroreflex-mediated effects (Ogoh et al. 2003) masked the magnitude of the impact of attenuated muscle afferent feedback on MAP (and maybe also ) and therefore FBF.

In an attempt to estimate the magnitude of the baroreflex-mediated compensatory effects on MAP during the fentanyl trial, we calculated MAP (FBF × leg vascular resistance) using FBF from the Ctrl trial and a leg vascular resistance from the fentanyl trial. These calculations suggest that MAP during the exercise with fentanyl (Fig. 2B) was at least 18% higher than would be expected as a result of the reduction in FBF. Although speculative, this analysis further emphasizes the substantial influence of muscle afferents on the regulation of MAP during exercise.

Exercise hyperpnoea and group III/IV muscle afferents

The relative contribution of thin fibre muscle afferents in the ventilatory control during exercise has been controversial. Numerous animal and human experiments isolating the effects of group III/IV muscle afferents indicate their critical involvement in exercise hyperpnoea (Kaufman & Forster, 1996). However, there has been little direct evidence emphasizing the necessity of sensory afferents in the regulation of the ventilatory response to dynamic whole body exercise (Galbo et al. 1987; Amann et al. 2010). The current study confirms these previous findings which utilized cycle exercise. In this context, it needs to be emphasized that the present results were obtained during an exercise modality which recruits a much smaller muscle mass, which evoked a substantially lower ventilatory response as compared to the previous studies (knee-extensor (current study): ∼20–40 l min⁻¹; bike (Amann et al. 2010): ∼40–150 l min⁻¹). Consequently, although the absolute reduction in was only small (1–3 l min⁻¹), alveolar hypoventilation caused significant and consistent CO₂ retention and a 3–4 Torr reduction in throughout exercise. Thus, the present study confirms our previous work (Amann et al. 2010) and emphasizes the important contribution of group III/IV muscle afferents in ventilatory control during voluntary exercise across the entire range of exercise hyperpnoea from only a small 1-fold to a much larger 15-fold increase in above rest.

Conclusion

Our results document the importance of continuous afferent feedback from working human skeletal muscle to achieve the appropriate haemodynamic and ultimately metabolic response to exercise. Therefore, despite attempts to compensate for the attenuated group III/IV muscle afferent effects by other mechanisms, we conclude that the relative contribution of muscle afferents to the haemodynamic response to exercise are clear and larger than traditionally accepted.

Glossary

Abbreviations

FBF

femoral blood flow

HR

heart rate

LVC

leg vascular conductance

MAP

mean arterial pressure

NA

noradrenaline

SV

stroke volume

Author contributions

All authors contributed to data collection/analysis and manuscript preparation. All authors approved the final version of the manuscript. The work was conducted at the VA Medical Center Salt Lake City, UT. The authors have no conflict of interest to disclose.