BACKGROUND MATERIAL FOR THURSDAY 18TH MAY

Transcription

1 Integrative human cardiovascular control Danish Cardiovascular Research Academy Ph.D. course The Panum Institute, University of Copenhagen Rigshospitalet May 15 20, 2017 BACKGROUND MATERIAL FOR THURSDAY 18TH MAY

2 J Physiol (2012) pp SYMPOSIUM REVIEW Inefficient functional sympatholysis is an overlooked cause of malperfusion in contracting skeletal muscle Bengt Saltin and Stefan P. Mortensen The Copenhagen Muscle Research Centre, Rigshospitalet, Denmark The Journal of Physiology Abstract Contracting skeletal muscle can overcome sympathetic vasoconstrictor activity (functional sympatholysis), which allows for a blood supply that matches the metabolic demand. This ability is thought to be mediated by locally released substances that modulate the effect of noradrenaline (NA) on the α-receptor. Tyramine induces local NA release and can be used in humans to investigate the underlying mechanisms and physiological importance of functional sympatholysis in the muscles of healthy and diseased individuals as well as the impact of the active muscles training status. In sedentary elderly men, functional sympatholysis and muscle blood flow are impaired compared to young men, but regular physical activity can prevent these age related impairments. In young subjects, two weeks of leg immobilization causes a reduced ability for functional sympatholysis, whereas the trained leg maintained this function. Patients with essential hypertension have impaired functional sympatholysis in the forearm, and reduced exercise hyperaemia in the leg, but this can be normalized by aerobic exercise training. The effect of physical activity on the local mechanisms that modulate sympathetic vasoconstriction is clear, but it remains uncertain which locally released substance(s) block the effect of NA and how this is accomplished. NO and ATP have been proposed as important inhibitors of NA mediated vasoconstriction and presently an inhibitory effect of ATP on NA signalling via P2 receptors appears most likely. (Received 16 July 2012; accepted after revision 3 September 2012; first published online 10 September 2012) Corresponding author B. Saltin: The Copenhagen Muscle Research Centre, Rigshospitalet, Section 7652, Blegdamsvej 9, DK-2100 Copenhagen, Denmark. bengt.saltin@rh.regionh.dk Abbreviations MSNA, muscle sympathetic nerve activity; NA, noradrenaline; L-NAME, N G -nitro-l-arginine methyl ester; L-NMMA, N G -monomethyl-l-arginine; ROS, reactive oxygen species; SNA, sympathetic nerve activity. Introduction The onset of exercise induces not only a very rapid increase in cardiac output but also a major redistribution of the blood flow. Sympathetic vasoconstrictor activity impairs perfusion of inactive muscles whereas active muscles abruptly increase their blood flow. Early studies providing evidence for this regulation were published in Bengt Saltin received his education from the medical school at the Karolinska Institute in Stockholm, where he also defended his doctoral thesis in Since then he has spent all his life performing human exercise related research, first in Stockholm, for some years in the USA and then primarily in Copenhagen, Denmark. This work ranges from patient groups and healthy sedentary people to elite-level atheletes of both sexes, focusing on skeletal muscle plasticity and regulation of hyperaemia and heart function. Stefan P. Mortensen (DMSc)earned his master s degree from the University of Copenhagen, was a PhD student in Niels H. Sechers laboratory and received post doctoral training with Bengt Saltin. He is currently leader of the cardiovascular group at the Centre of Inflammation and Metabolism at Rigshospitalet, Demark. His main research interest is in the regulation of the cardiovascular system during exercise and how cardiovascular function is altered in disease states. This report was presented at The Journal of Physiology Symposium on Blood flow regulation: from rest to maximal exercise, which took place at the Main Meeting of The Physiological Society, Edinburgh, UK on 3 July It was commissioned by the Editorial Board and reflects the views of the authors. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013 DOI: /jphysiol

3 6270 B. Saltin and S. P. Mortensen J Physiol and 1962: Rein (1937) showed that an elevated vasomotor outflow caused a reduction in the blood flow to resting muscles, and Remensnyder et al. (1962) confirmed that vasoconstriction is abolished in contracting muscle and termed this phenomenon functional sympatholysis, i.e. the vasoconstrictor effect of the noradrenaline (NA) release was ineffective or blocked. At the time no precise definition was given of functional sympatholysis, but a definition could be that it is a direct action of one or more exercise-produced compound(s) that block the vasoconstrictive effect of NA via its receptor on smooth muscles. In contrast, vasodilatation is the action of a multitude of factors affecting the tonus of smooth muscles including the endothelial cells and upstream vasodilatation (Clifford & Hellsten, 2004; Hellsten et al. 2012). In the following, some old and newer data will be presented to demonstrate the crucial significance of functional sympatholysis for achieving a blood flow that matches the energy demand of human skeletal muscles. However, the reader should be reminded that during intense exercise involving large muscle groups, functional sympatholysis has to be overcome to preserve blood pressure as peak local muscle blood flow can reach ml kg 1 min 1, which would require a cardiac output close to 100 l min 1 (Saltin, 2007). and oxygen delivery to contracting muscles in exercising humans. Tyramine. A limitation to the study of functional sympatholysis in humans has been the lack of interventions to either manipulate the sympathetic outflow or block its action at the receptor site. Tyramine can elevate the NA release from sympathetic nerve terminals and it was first used almost a century ago in studies of the circulation (Tainter, 1929). Infusion of tyramine caused a pressure response in a wide range of species. In a series of experiments in the following decades some of the complex actions of tyramine were evaluated, which revealed a strong α-receptor effect, butalso thata very high dose stimulated the β-receptors, which resulted in some vasodilatation (for references see Brandão et al. (1978) and Trendelenburg (1972)). Compared to NA, tyramine elicited a slightly slower response, which was thought to be attributed to the time for tyramine to act on the sympathetic nerves for the NA release to become increased. Although one study showed that sympatectomy abolished the tyramine effect (Frewin & Whelan, 1968), the belief is that tyramine primarily has its effect on the nerve terminal (Eisenach et al. 2002). Several of the early studies were reproduced in the 1960s in the human forearm (Cohn, Functional sympatholysis Early studies. The capacity to overcome even very high sympathetic nerve activity (SNA) is demonstrated in studies with dynamic one-legged knee-extensor exercise, where a very marked increase in sympathetic activity accomplished by adding forearm ischemic static exercise, did not alter blood flow to the contracting knee extensor muscles, in spite of the 2- to 4-fold higher muscle sympathetic nerve activity (MSNA) (Fig. 1A)(Strange, 1999). Similar findings were reported when SNA was increased by enlarging the muscle mass involved in the exercise (Fig. 1B). One-legged knee-extensor exercise was performed at a given high submaximal power output with measurements of blood flow and NA spillover. When the active muscle mass was increased by adding exercise with the arms and the other leg, NA spillover increased 4-fold above the control exercise, but leg blood flow was unaltered (Savard et al. 1989). In contrast to the study with static forearm contractions and ischaemia, blood pressure was unchanged and so was vascular conductance (whereas the conductance was lowered with the forearm static contraction as it also caused an elevation in blood pressure). These data and other similar observations in the literature (Joyner et al. 1992; Rosenmeier et al. 2004) support the notion of a powerful role of functional sympatholysis in obtaining an appropriate and well-matched blood flow Figure 1. Muscle sympathetic nerve activity during leg exercise with and without static handgrip exercise (SHG) and forearm ischaemia (F I) (A) and noradrenaline spillover and blood flow in the exercising leg with and without additional active muscle groups (B) Modified from Strange 1999 (A) and Savard et al (B). Licence Number C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

4 J Physiol Muscle blood flow in healthy ageing ; Frewin & Whelan, 1968), but for a long period the regulation of skeletal muscle microcirculatory perfusion became a question of vasodilatation rather than functional sympatholysis. Mediators of functional sympatholysis in humans. To determine the mechanisms of functional sympatholysis, experiments have been conducted to identlify compounds with an effect that could mimic functional sympatholysis, i.e. block the effect of local tyramine infusion or systemic sympathetic stimulation in a similar manner as exercise blocksthevasoconstrictiveeffectofna. Nitric oxide. Nitric oxide (NO) has been suggested to block the α-receptors in rat muscles (Thomas & Victor, 1998; Thomas et al. 2003). Some support for a role of NO in humans can also be found in children with Duchenne muscular dystrophy who are n-nos deficient and have impaired functional sympatholysis during exercise (Sander et al. 2000). Inhibition of systemic NO formation with N G -nitro-l-arginine methyl ester (L-NAME) has been shown to blunt the modulation of sympathetic vasoconstrictor activity in the exercising forearm (Chavoshan et al. 2002), but local inhibition of NO formation with N G -monomethyl-l-arginine (L-NMMA) does not have the same effect (Dinenno & Joyner, 2003). Impaired functional sympatholysis has also been linked to increased levels of reactive oxygen species (ROS) (Fadel et al. 2012). Two important observations appear to rule out an essential role of NO in blocking sympathetic vasoconstriction. Firstly, exogenous NO stimulation of NO availability does not blunt sympathetic vasoconstriction (Rosenmeier et al. 2003). Secondly, inhibition of NO formation does not reduce exercise hyperaemia at the early onset or the steady-state levels even during maximal knee extensions where the SNA is markedly increase (Fig. 2) (Rådegran & Saltin, 1999; Savard et al. 1989). Although these observations point to a non-obligatory role of NO, it could act in synergy with other substances such as prostacyclin (Dinenno & Joyner, 2004) to mediate functional sympatholysis in a similar fashion as the combined role of these two substances in skeletal muscle vasodilatation (Mortensen et al. 2007). ATP. Studies by Rosenmeier et al. (2004, 2008) employed infusion of tyramine and various purines and pyramines as well as exercise to show that UTP and ATP produced a response, which mimicked functional sympatholysis during exercise. In these studies the focus was on nucleotides in plasma, which implies a role of endothelial P2Y receptors to mediate their effect, because arterial ATP does not appear to cross into the interstitial space of the muscle (Mortensen et al. 2009b). We recently found that 2 weeks of leg immobilization markedly lower interstitial ATP concentrations during exercise (S. P. Mortensen, J. H. Svendsen, Y. Hellsten, N. H. Secher & B. Saltin, in review) in association with an impaired functional sympatholysis (Fig. 3A; Mortensen et al. 2012). This opens up the possibility that interstitial ATP interacts directly with the α-receptor possibly via P2Y receptors or via endothelial P2Y receptors (Mortensen et al. 2009a). A likely scenario could be that luminal and interstitial ATP operate synergistically via P2 receptors, which are also markedly affected by the level of physical activity (Fig. 4; Mortensen et al. 2012). The origin of plasma ATP has been proposed to be endothelial and red blood cells (Mortensen et al. 2011), whereas the skeletal muscle is the main source of interstitial ATP (Tu et al. 2010). Functional sympatholysis as a cause of malperfusion with ageing and in disease Ageing. Wahren et al. (1974) first described that leg blood flow during cycling at a given submaximal power output was lower in middle aged individuals as compared with young individuals. More than 25 years later, Proctor et al reported that ageing was associated with an increase in sympathetic vasoconstrictor activity during exercise. Both of these finding have been confirmed in later studies (Koch et al. 2003; Dinenno et al. 2005; Kirby et al. 2011). A link between the reduced exercise hyperaemia and impaired functional sympatholysis remains to be established, but we have recently observed that chronically endurance trained elderly can maintain an intact functional sympatholysis and sufficient blood flow to maintain aerobic metabolism (Fig. 3A; S.P.Mortensen,M.Nyberg,K.Winding&B. Saltin, in review). In contrast, a similar leg blood flow in chronically sedentary elderly was associated with a lower aerobic metabolism, increased lactate release and impaired Figure 2. Leg vascular conductance (VC) at rest and during submaximal and maximal exercise with and without inhibition of nitric oxide formation (L-NMMA) Modified from Rådegran & Saltin (1999). C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

5 6272 B. Saltin and S. P. Mortensen J Physiol functional sympatholysis. These observations illustrate the need for local determination of muscle perfusion and metabolism to determine the physiological importance of functional sympatholysis and separate the positive effect of long term exercise training on limb blood flow from the age related impairments in blood flow regulation. Despite of a reduced endothelial function (Taddei et al. 1995) and sympatholysis with ageing, the vasodilatory responsiveness and the sympatholytic effect of arterially infused ATP appear to be maintained in the forearm (Kirby et al. 2010, 2011), suggesting that a role of plasma ATP would have to be caused by reduced levels of plasma ATP. Indeed, interstitial ATP concentrations were reduced in the sedentary elderly compared to the active elderly (Fig.5;S.P.Mortensen,M.Nyberg,K.Winding&B. Saltin, in review). Also, skeletal muscle hyperaemia was not increased when NO bioavailability was increased by acute antioxidant infusion, suggesting that NO is not involved in the lowering of exercise hyperaemia in the ageing leg (M. Nyberg, M. Hellsten & S. P. Mortensen, in review). Figure 3. Percentage change in leg blood flow and vascular conductance when tyramine was infused during one-legged knee extensor exercise with a control, detrained and trained leg (A) and in young, sedentary elderly and trained elderly (right panel) A, Different from control leg; P < 0.05, #different from immobilized leg. B, P < 0.05, different from young men, P < 0.05; different from sedentary elderly men, P < Modified from Mortensen et al. 2012a, 2012b. Figure 4. Skeletal muscle interstitial P2Y2 receptor content and localization A, P2Y2 receptor content in vastus lateralis muscle before and after 14 days of immobilization of one leg and 5 weeks of exercise training of the other leg. Different from the control leg, P < 0.05; #different from the immobilized leg, P < B, immunohistochemical localization of purinergic P2Y 2 receptors in human skeletal muscle. Positive staining for P2Y 2 and endothelium superimposed. P2Y 2 purinergic receptors were evident in endothelial cells of capillaries and microvessels (white arrow), and in vascular smooth muscle cells (white arrowhead). Modified from Mortensen et al. 2009a, C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

6 J Physiol Muscle blood flow in healthy ageing 6273 Disease. Several disease states have been characterized by a reduced blood flow during exercise (Esposito et al. 2011; Nyberg et al. 2012) and increased levels of sympathetic vasoconstrictor activity (Parati & Esler, 2012), but the role of functional sympatholysis in this apparent malperfusion remains largely unexplored. Recent evidence suggests that functional sympatholysis is impaired in the exercising forearm of middle-aged men and women with essential hypertension (Vongpatanasin et al. 2011). Surprisingly, the forearm blood flow and vascular conductance during exercise were not different between the normotensive and hypertensive subjects, especially considering that the hypertensive subjects also had a higher MSNA response to exercise (Vongpatanasin et al. 2011). In the exercising leg, blood flow is lower in sedentary hypertensive subjects compared to normotensive subjects matched for activity and age (Nyberg et al. 2012). Moreover, 8 weeks of high intensity exercise normalized the blood flow response to exercise (Nyberg et al. 2012). In contrast to the impaired functional sympatholysis in hypertensive subjects, tyramine did not lower moderate intensity exercise hyperaemia in type II diabetics, suggesting that functional sympatholysis may be intact in this patient group (Thaning et al. 2011). However, these patients also had a normal vasodilator response to ACh and it is therefore possible that functional sympatholysis is affected in type 2 diabetics when endothelial dysfunction is also present. With regard to the role of plasma ATP, the sympatholytic properties of ATP appear to be reduced in type 2 diabetics compared to healthy controls and the vasodilator effect of arterially infused ATP is also reduced (Thaning et al. 2010). Taken together, these observations suggest that plasma ATP may not be obligatory in mediating functional sympatholysis during exercise, especially given that in vitro observations suggest that type 2 diabetics also have impaired ATP release from red blood cells (Sprague et al. 2011). Summary and perspective Since the observation by Remensnyder et al. (1962) of the exercise mediated blunting of an elevated sympathetic nerve activity, which they named functional sympatholysis, ample support is now available demonstrating its role in regulating muscle hyperaemia also in exercising humans. Although not yet fully documented, recent findings indicate that ATP may be a principal mediator of this effect via its binding to a P2 receptor (Fig. 6) and both luminal and interstial ATP are in play. This proposal is based on data from studies on human leg muscles (Rosenmeier et al. 2004, 2008; Mortensen et al. 2012). A major future task is to bridge the gap in our knowledge of similarities and dissimilarities in the regulation of the microcirculatory blood flow in legs and arms (Newcomer et al. 2004; Thijssen et al. 2011). This also relates to functional sympatholysis. A loss of an efficient functional sympatholysis contributes to the reduced muscle perfusion observed in ageing muscles, which is less pronounced in physically active elderly, indicating that lack of muscle usage in older individuals is the primary cause rather than ageing per se. In the hamster model, both ageing and SNA impair upstream vasodilatation (Bearden et al. 2004; Haug & Segal, 2005) and it is therefore tempting to propose that functional sympatholysis may involve substances that Figure 5. Muscle interstitial ATP concentrations at rest and during exercise at 12 W and 45% of maximal workload in young, sedentary elderly and active elderly men Different from baseline conditions, P < 0.05; different from young men (same condition), P < 0.05; different from sedentary elderly men (same condition), P < Modified from Mortensen et al. 2012b. Figure 6. Simplified model of functional sympatholysis, tyramine action and potential role of ATP Luminal and interstial ATP can block the vasoconstrictor effect of noradrenaline (NA) released from sympathetic nerve terminals via P2 receptors. Sympathetic vasoconstrctor activity reduces upstream vasodilation by impairing endothelial signaling and functional sympatholysis may also act by restoring upstream vasodilatation (blue arrows). In humans, NA release can be stimulated by arterial infusion of tyramine. See text for further details. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

7 6274 B. Saltin and S. P. Mortensen J Physiol counteract the vasoconstrictor effect of SNA by restoring endothelial cell-to-cell signalling via gap junctions. If this is the case, the primary cause for a lowering of the muscle blood flow in ageing is lack of dilatation of the small feeding arterioles and poor functional sympatholysis, and not an inefficient ordinary vasodilatation. References Bearden SE, Payne GW, Chisty A & Segal SS (2004). Arteriolar network architecture and vasomotor function with ageing in mouse gluteus maximus muscle. JPhysiol561, Brandão F, Monteiro JG & Osswald W (1978). Differences in the metabolic fate of noradrenaline released by electrical stimulation or by tyramine. Naunyn Schmiedebergs Arch Pharmacol 305, Chavoshan B, Sander M, Sybert TE, Hansen J, Victor RG & Thomas GD (2002). Nitric oxide-dependent modulation of sympathetic neural control of oxygenation in exercising human skeletal muscle. JPhysiol540, Clifford PS & Hellsten Y (2004). Vasodilatory mechanisms in contracting skeletal muscle. JApplPhysiol97, Cohn JN (1965). Comparative cardiovascular effects of tyramine, ephedrine, and norepinephrine in man. Circ Res 16, Dinenno FA & Joyner MJ (2003). Blunted sympathetic vasoconstriction in contracting skeletal muscle of healthy humans: is nitric oxide obligatory? JPhysiol553, Dinenno FA & Joyner MJ (2004). Combined NO and PG inhibition augments α-adrenergic vasoconstriction in contracting human skeletal muscle. Am J Physiol Heart Circ Physiol 287, H2576 H2584. Dinenno FA, Masuki S & Joyner MJ (2005). Impaired modulation of sympathetic α-adrenergic vasoconstriction in contracting forearm muscle of ageing men. JPhysiol567, Eisenach JH, Clark ES, Charkoudian N, Dinenno FA, Atkinson JLD, Fealey RD, Dietz NM & Joyner MJ (2002). Effects of chronic sympathectomy on vascular function in the human forearm. JApplPhysiol92, Esposito F, Reese V, Shabetai R, Wagner PD & Richardson RS (2011). Isolated quadriceps training increases maximal exercise capacity in chronic heart failure: the role of skeletal muscle convective and diffusive oxygen transport. JAmColl Cardiol 58, Fadel PJ, Farias M, Gallagher KM, Wang Z & Thomas GD (2012). Oxidative stress and enhanced sympathetic vasoconstriction in contracting muscles of nitrate tolerant rats and humans. JPhysiol590, Frewin DB & Whelan RF (1968). The action of ephedrine on forearm blood vessels in man. Aust J Exp Biol Med Sci 46, Haug SJ & Segal SS (2005). Sympathetic neural inhibition of conducted vasodilatation along hamster feed arteries: complementary effects of α 1 -andα 2 -adrenoreceptor activation. JPhysiol563, Hellsten Y, Nyberg M & Mortensen SP (2012). Contribution of intravascular versus interstitial purines and nitric oxide in the regulation of exercise hyperaemia in humans. JPhysiol 590, Joyner MJ, Nauss LA, Warner MA & Warner DO (1992). Sympathetic modulation of blood flow and O 2 uptake in rhythmically contracting human forearm muscles. Am J Physiol Heart Circ Physiol 263, H1078 H1083. Kirby BS, Crecelius AR, Voyles WF & Dinenno FA (2010). Vasodilatory responsiveness to adenosine triphosphate in ageing humans. JPhysiol588, Kirby BS, Crecelius AR, Voyles WF & Dinenno FA (2011). Modulation of postjunctional α-adrenergic vasoconstriction during exercise and exogenous ATP infusions in ageing humans. JPhysiol589, Koch DW, Leuenberger UA & Proctor DN (2003). Augmented leg vasoconstriction in dynamically exercising older men during acute sympathetic stimulation. JPhysiol551, Mortensen SP, González-Alonso J, DamsgaardR, Saltin B & Hellsten Y (2007). Inhibition of nitric oxide and prostaglandins, but not endothelial-derived hyperpolarizing factors, reduces blood flow and aerobic energy turnover in the exercising human leg. JPhysiol581, Mortensen SP, Gonzalez-Alonso J, Bune LT, Saltin B, Pilegaard H & Hellsten Y (2009a). ATP-induced vasodilation and purinergic receptors in the human leg: roles of nitric oxide, prostaglandins, and adenosine. Am J Physiol Regul Integr Comp Physiol 296, R1140 R1148. Mortensen SP, Gonzalez-Alonso J, Nielsen JJ, Saltin B & Hellsten Y (2009b). Muscle interstitial ATP and norepinephrine concentrations in the human leg during exercise and ATP infusion. JApplPhysiol107, Mortensen SP, Mørkeberg J, Thaning P, Hellsten Y & Saltin B (2012a). Two weeks of muscle immobilization impairs functional sympatholysis, but increases exercise hyperemia and the vasodilatory responsiveness to infused ATP. Am J Physiol Heart Circ Physiol 302, H2074 H2082. Mortensen SP, Nyberg M, Winding K. & Saltin B (2012b). Lifelong physical activity preserves functional sympatholysis and purinergic signaling in the human leg JPhysiol jphysiol ; published ahead of print September 10, doi: /jphysiol Mortensen SP, Thaning P, Nyberg M, Saltin B & Hellsten Y (2011). Local release of ATP into the arterial inflow and venous drainage of human skeletal muscle: insight from ATP determination with the intravascular microdialysis technique. JPhysiol589, Newcomer SC, Leuenberger UA, Hogeman CS, Handly BD & Proctor DN (2004). Different vasodilator responses of human arms and legs. JPhysiol556, Nyberg M, Jensen LG, Thaning P, Hellsten Y & Mortensen SP (2012). Role of nitric oxide and prostanoids in the regulation of leg blood flow and blood pressure in humans with essential hypertension: effect of high-intensity aerobic training. JPhysiol590, Parati G & Esler M (2012). The human sympathetic nervous system: its relevance in hypertension and heart failure. Eur Heart J 33, C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

8 J Physiol Muscle blood flow in healthy ageing 6275 Proctor DN, Shen PH, Dietz NM, Eickhoff TJ, Lawler LA, Ebersold EJ, Loeffler DL & Joyner MJ (1998). Reduced leg blood flow during dynamic exercise in older endurance-trained men. JApplPhysiol85, Rådegran G & Saltin B (1999). Nitric oxide in the regulation of vasomotor tone in human skeletal muscle. Am J Physiol Heart Circ Physiol 276, H1951 H1960. Rein HSM (1937). Die lokale Stoffwechseleinschrankung bei reflektoriseh-nervoser Durehblutungdrosselung. Arch Ges Physiol 239, 464. Remensnyder J, Mitchell JH & Sarnoff SJ (1962). Functional sympatholysis during muscular activity. Observations on influence of carotid sinus on oxygen uptake. Circ Res 11, Rosenmeier JB, Hansen J & González-Alonso J (2004). Circulating ATP-induced vasodilatation overrides sympathetic vasoconstrictor activity in human skeletal muscle. JPhysiol558, Rosenmeier JB, Yegutkin GG & Gonzalez-Alonso J (2008). Activation of ATP/UTP selective receptors increase blood flow and blunt sympathetic vasoconstriction in human skeletal muscle. JPhysiol586, Rosenmeier JB, Fritzlar SJ, Dinenno FA & Joyner MJ (2003). Exogenous NO administration and α-adrenergic vasoconstriction in human limbs. JApplPhysiol95, Saltin B (2007). Exercise hyperaemia: magnitude and aspects on regulation in humans. JPhysiol583, Sander M, Chavoshan B, Harris SA, Iannaccone ST, Stull JT, Thomas GD & Victor RG (2000). Functional muscle ischemia in neuronal nitric oxide synthase-deficient skeletal muscle of children with Duchenne muscular dystrophy. Proc Natl Acad Sci U S A 97, SavardGK,RichterEA,StrangeS,KiensB,ChristensenNJ& Saltin B (1989). Norepinephrine spillover from skeletal muscle during exercise in humans: role of muscle mass. Am J Physiol Heart Circ Physiol 257, H1812 H1818. Sprague RS, Bowles EA, Achilleus D, Stephenson AH, Ellis CG & Ellsworth ML (2011). A selective phosphodiesterase 3 inhibitor rescues low PO 2 -induced ATP release from erythrocytes of humans with type 2 diabetes: implication for vascular control. Am J Physiol Heart Circ Physiol 301, H2466 H2472. Strange S (1999). Cardiovascular control during concomitant dynamic leg exercise and static arm exercise in humans. JPhysiol514, Taddei S, Virdis A, Mattei P, Ghiadoni L, Gennari A, Fasolo CB, Sudano I & Salvetti A (1995). Aging and endothelial function in normotensive subjects and patients with essential hypertension. Circulation 91, Tainter M (1929). The actions of tyramine on the circulation and smooth muscle. JPharmacol30, Thaning P, Bune LT, Zaar M, Saltin B & Rosenmeier JB (2011). Functional sympatholysis during exercise in patients with type2diabeteswithintactresponsetoacetylcholine. Diabetes Care 34, Thaning P, Bune LT, Hellsten Y, Pilegaard H, Saltin B & Rosenmeier JB (2010). Attenuated purinergic receptor function in patients with type 2 diabetes. Diabetes 59, Thijssen DHJ, Rowley N, Padilla J, Simmons GH, Laughlin MH, Whyte G, Cable NT & Green DJ (2011). Relationship between upper and lower limb conduit artery vasodilator function in humans. JApplPhysiol111, Thomas GD, Shaul PW, Yuhanna IS, Froehner SC & Adams ME (2003). Vasomodulation by skeletal muscle-derived nitric oxide requires α-syntrophin-mediated sarcolemmal localization of neuronal nitric oxide synthase. Circ Res 92, Thomas GD & Victor RG (1998). Nitric oxide mediates contraction-induced attenuation of sympathetic vasoconstriction in rat skeletal muscle. JPhysiol506, Trendelenburg U (1972). Classification of sympathomimetic amines. In Handbook of Experimental Pharmacology,ed. Blaschko H & Muscholl E, pp Springer, Berlin. Tu J, Le G & Ballard HJ (2010). Involvement of the cystic fibrosis transmembrane conductance regulator in the acidosis-induced efflux of ATP from rat skeletal muscle. JPhysiol588, Vongpatanasin W, Wang Z, Arbique D, Arbique G, Adams-Huet B, Mitchell JH, Victor RG & Thomas GD (2011). Functional sympatholysis is impaired in hypertensive humans. JPhysiol589, Wahren J, Saltin B, Jorfeldt L & Pernow B (1974). Influence of age on the local circulatory adaptation to leg exercise. Scand J Clin Lab Invest 33, Acknowledgements B.S. is supported by the Collstrop Foundation. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on February 2, 2013

9 J Physiol (2012) pp SYMPOSIUM REVIEW Vasodilator interactions in skeletal muscle blood flow regulation Y. Hellsten 1,M.Nyberg 1,L.G.Jensen 1 and S. P. Mortensen 2 1 Department of Exercise and Sport Sciences, University of Copenhagen, Copenhagen, Denmark 2 Centre of Inflammation and Metabolism, Rigshospitalet, Copenhagen, Denmark The Journal of Physiology Abstract During exercise, oxygen delivery to skeletal muscle is elevated to meet the increased oxygen demand. The increase in blood flow to skeletal muscle is achieved by vasodilators formed locally in the muscle tissue, either on the intraluminal or on the extraluminal side of the blood vessels. A number of vasodilators have been shown to bring about this increase in blood flow and, importantly, interactions between these compounds seem to be essential for the precise regulation of blood flow. Two compounds stand out as central in these vasodilator interactions: nitric oxide (NO) and prostacyclin. These two vasodilators are both stimulated by several compounds, e.g. adenosine, ATP, acetylcholine and bradykinin, and are affected by mechanically induced signals, such as shear stress. NO and prostacyclin have also been shown to interact in a redundant manner where one system can take over when formation of the other is compromised. Although numerous studies have examined the role of single and multiple pharmacological inhibition of different vasodilator systems, and important vasodilators and interactions have been identified, a large part of the exercise hyperaemic response remains unexplained. It is plausible that this remaining hyperaemia may be explained by camp- and cgmp-independent smooth muscle relaxation, such as effects of endothelial derived hyperpolarization factors (EDHFs) or through metabolic modulation of sympathetic effects. The nature and role of EDHF as well as potential novel mechanisms in muscle blood flow regulation remain to be further explored to fully elucidate the regulation of exercise hyperaemia. (Submitted 11 July 2012; accepted after revision 10 September 2012; first published online 17 September 2012) Corresponding author Y. Hellsten: Department of Exercise and Sport Sciences, Division of Integrated Cardiovascular Physiology, University of Copenhagen, Universitetsparken 13, DK-2100 Copenhagen, Denmark. yhellsten@ifi.ku.dk Abbreviations ACh, acetylcholine; AA, arachidonic acid; BKR, bradykinin receptor 2; CaM, calmodulin; CaMK, calmodulin kinase; camp, cyclic adenosine monophosphate; cgmp, cyclic guanosine monophosphate, COX, cyclooxygenase; CYP 2C9, cytochrome P450 2C9; EET, eicosatrienoic acid; EDHF, endothelial derived hyperpolarizing factor; HR, histamine receptor; PGI 2, prostacyclin; PKA, protein kinase A; PKC, protein kinase C; P1, purinergic receptor 1; P2, purinergic receptor 2; SMC, smooth muscle cell. Ylva Hellsten is head of the cardiovascular research group at the Department of Exercise and Sport Sciences, Section for Integrated Physiology, University of Copenhagen. The research group investigates the regulation of skeletal muscle blood flow and skeletal muscle angiogenesis in health and cardiovascular disease. Senior researcher Stefan P. Mortensen, DMSci, is leader of the cardiovascular group at the Centre of Inflammation and Metabolism at Rigshospitalet. He earned his master s degree from the University of Copenhagen and received post-doctoral training with Professor Bengt Saltin in Copenhagen. His main research interest is cardiovascular regulation during exercise and alterations in disease states. This report was presented at The Journal of Physiology Symposium on Blood flow regulation: from rest to maximal exercise, which took place at the Main Meeting of The Physiological Society, Edinburgh, UK on 3 July It was commissioned by the Editorial Board and reflects the views of the authors. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012 DOI: /jphysiol

10 6298 Y. Hellsten and others J Physiol Introduction Blood flow to skeletal muscle is highly dynamic and increases markedly with exercise at a rate closely related to the oxygen demand of the muscle (Andersen & Saltin, 1985). Overall, muscle blood flow is regulated through a balance between, on the one hand, sympathetic activity and vasoconstrictors and, on the other hand, vasodilators and compounds modulating the effect of sympathetic activity. These vasodilating compounds are formed locally in the skeletal muscle tissue and are released from endothelial cells, red blood cells and skeletal muscle cells as a result of signals primarily related to the balance between oxygen delivery and demand. Several vasodilators, including nitric oxide (NO), prostacyclin, ATP, adenosine, potassium and compounds associated with the endothelium derived hyperpolarizing factor (EDHF) concept, such as 11,12-eicosatrienoic acid (11,12-EET), have been proposed to be of importance for muscle blood flow regulation. For review on this topic see Clifford & Hellsten (2004) and Sarelius & Pohl (2010). Evidence for the role of these vasodilators in exercise hyperaemia stems from studies showing that the vasodilators are formed in exercising muscle and from studies using pharmacological interventions to either inhibit or promote the vasodilator systems. None of the proposed vasodilators seem to operate independently or to be essential for reaching adequate blood flow during exercise, but they show a close interaction with other vasodilator systems. Vasodilator interactions may serve two purposes, where one is a redundancy mechanism whereby one vasodilator can take over when the formation of another vasodilator is impaired, and the other is activation of other vasodilator systems. The redundancy interaction may occur either chemically, by direct interactions between the vasodilator systems, or be functional and coupled to the demand for oxygen. Redundancy is a physiologically important concept as it can secure adequate oxygen supply despite impairments in vasodilator function. It is important to keep in mind that functional redundancy only becomes apparent in experimental settings when there is a demand for oxygen in the tissue, such as during exercise or hypoxia, whereas it is lacking in in vitro set-ups and experiments utilizing infusion of vasodilators. The other kind of vasodilator interaction serves to promote the formation of one or several other vasodilating systems, thereby potentially enhancing the vasodilator effect. NO and prostacyclin appear to be central in both of these interactions as they share a redundancy interaction and as they both are activated by multiple compounds and mechanical signals. It should be emphasized that several other compounds thanmentionedinthisreviewhavebeenproposedto contribute to exercise hyperaemia, e.g. potassium and lactate. Due to restrictions in length of this review we have chosen to discuss only selected compounds and their interactions and with a focus on human studies. Functional role of NO, prostanoids and EDHF in exercise hyperaemia In 1969 prostanoids were proposed to be involved in muscle blood flow regulation based on the findings that infusion of prostanoids into the brachial artery increased blood flow (Bevegård & Orö, 1969). A role for prostanoids in exercise hyperaemia was later supported by the findings that both plasma (Wilson & Kapoor, 1993) and interstitial (Frandsen et al. 2000) prostacyclin and prostaglandin E 2 concentrations were increased during muscle contractions in the forearm and leg, respectively. However, infusion of cyclooxygenase (COX) inhibitors to inhibit the formation of prostanoids, has shown no effect on blood flow at rest or during exercise in the human forarm (Shoemaker et al. 1996) or leg (Fig. 1) (Mortensen et al. 2007; Schrage et al. 2010). In the late 1980s, Vallance and co-workers blocked NO synthase (NOS) by infusion of N G -monomethyl-l-arginine (L-NMMA) and showed a 50% reduction in resting forearm blood flow (Vallance et al. 1989). The importance of NO for resting blood flow, blood flow in recovery from exercise, and blood flow during passive movement has since been widely confirmed in the leg (Rådegran & Saltin, 1999; Mortensen et al. 2009b; Heinonen et al. 2011) and in the forearm (Panza et al. 1993; Gilligan et al. 1994; Dyke et al. 1995). During exercise, however, inhibition of NO formation has been shown not to reduce blood flow to the leg (Rådegran & Saltin, 1999; Bradley et al. 1999; Frandsen et al. 2001; Kingwell et al. 2002; Heinonen et al. 2011), whereas a transient effect has been observed in the forearm (Schrage, 2004). Thus, at least in the leg, neither NO nor prostanoids appear to be obligatory for exercise hyperaemia (Fig. 1). However, experiments in which the synthesis of NO and prostanoids have been inhibited simultaneously have demonstrated a clear reduction in leg blood flow during exercise (Boushel et al. 2002; Mortensen et al. 2007; 2009b; Heinonen et al. 2011). The observation that single inhibition of a system has no effect on exercise hyperaemia whereas combined inhibition markedly lowers blood flow suggests that there is a compensatory formation of the other vasodilator so that adequate blood flow is achieved. Direct interactions between the vasodilator systems may explain this redundancy, as described below. EDHF is a concept derived from the observation that acetylcholine induces hyperpolarization of smooth muscle cells in the presence of NOS and COX inhibitors (Busse et al. 2002). The EDHF concept, which has not yet been fully elucidated, entails multiple compounds and the identity of these compounds varies according to tissue C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

11 J Physiol Vasodilator interactions in skeletal muscle blood flow regulation 6299 and blood vessel type (Busse et al. 2002). In coronary and skeletal muscle resistance arteries, the product of CYP 2C9, 11,12-eicosatrienoic acid (11,12-EET), as well as other eicosatrienoic acids (EETs), such as 8,9- and 14,15-EETs, have been identified as EDHFs (Fisslthaler et al. 1999; Bolz et al. 2000). In the human forearm, bradykinin has been shown to elevate blood flow during NOS and COX blockade, indicating the presence of an EDHF mediated mechanism (Halcox et al. 2001). In the human leg, single blockade of CYP 2C9 by arterial infusion of sulphaphenazole has no effect on exercise hyperaemia but combined inhibition of NO synthesis and CYP 2C9 lowers blood flow by 15% (Fig. 1) (Hillig et al. 2003). This finding suggests an interaction between these two vasodilator systems, similar to the redundancy interaction observed between NO synthase and COX. Interestingly, when EDHF blockade by infusion of the non-specific potassium channel blocker tetraammonium chloride (TEA) is added to combined inhibition of NO synthase and COX, exercise hyperaemia is not reduced beyond that of NO synthase and COX inhibition combined (Mortensen et al. 2007). The immediate interpretation could be that EDHF induced vasodilatation cannot compensate for the impaired NO and prostanoid systems, but an equally possible explanation is that TEA infusion is not sufficiently specific to examine an EDHF effect. Figure 1. Change in exercise hyperaemia in the leg during inhibition of nitric oxide, prostaglandins, EDHF and the purinergic P1 receptor in young healthy male subjects A, relative change in leg exercise hyperaemia with single or combined inhibition of nitric oxide synthase (NOS) and the endothelial derived hyperpolarizing factor (EDHF) cytochrome P450 2C9. B, relative change in leg exercise hyperaemia with single, double or triple blockade of NOS, EDHF and cyclooxygenase (COX). C, relative change in leg exercise hyperaemia with single, double or triple blockade of NOS, COX and the adenosine P1 receptor. In all experiments the exercise performed was single leg knee extensor exercise. Inhibition was achieved by arterial or venous infusion: NOS inhibition by N G -monomethyl-l-arginine (L-NMMA) or N G -nitroarginine methyl ester (L-NAME), COX inhibition by indomethacin, CYP 2 C9 by sulfaphenazole, EDHF by tetraammoniumchloride, and P i receptor blockade by theophylline. Blood flow was determined by either thermodilution or ultrasound Doppler technique. Adapted from Frandsen et al. (2000), Hillig et al. (2002), Mortensen et al. (2007) and Mortensen et al. (2009). Stimulators of nitric oxide and prostanoid formation in endothelial cells NO produced from enos and prostacyclin produced by the COX pathway play critical roles in normal vascular biology and pathophysiology and regulate vascular conductance, platelet aggregation, angiogenesis and vascular smooth muscle proliferation (Dudzinski & Michel, 2007; Félétou et al. 2011). enos is activated by increases in intracellular Ca 2+ with concurrent binding of calmodulin (CaM) to the enzyme (Nathan & Xie, 1994) and by protein phosphorylation at several sites (Fleming & Busse, 2003; Mount et al. 2007). The formation of prostacyclin is stimulated by increases in intracellular Ca 2+ in endothelial cells that lead to liberation of arachidonic acid and activation of the COX pathway (Fig. 2) (Carter et al. 1988; Ray & Marshall, 2006; Domeier & Segal, 2007). Numerous chemical and mechanical stimuli including ATP, adenosine, ACh, insulin, bradykinin, histamine, thrombin, stretch and shear stress have been shown to activate enos through increases in Ca 2+ and/or phosphorylation status (de Wit et al. 1997; Fleming & Busse, 2003; Ray & Marshall, 2006; Domeier & Segal, 2007; Dudzinski & Michel, 2007; da Silva et al. 2009; Nyberg et al. 2010; Raqeebet al. 2011). Similarly, prostacyclin formation has been shown to be increased by several of the same stimuli, e.g. ACh, ATP, adenosine, bradykinin, histamine C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

12 6300 Y. Hellsten and others J Physiol and shear stress (Baenziger et al. 1980; Grabowski et al. 1985; Carter et al. 1988; Koller et al. 1994; Ray & Marshall, 2006; Domeier & Segal, 2007; Nyberg et al. 2010). The multiple stimulators of enos and the COX pathway highlight the importance of these vasoactive systems for vascular function and illustrate a dynamic control of NO and prostanoid bioactivity. The advantage of this design is that activation can occur even if some activation pathways are weak. On the other hand, impairments in these central vasodlator systems, as can occur in cardiovascular disease (Vanhoutte et al. 2009) can have a large impact on vascular function. Figure 2. Schematic illustration of vasodilator interactions in skeletal muscle arterioles A, illustration of a simplified view of endothelium-dependent induction of vasodilatation. Vasodilator agonist, such as adenosine or acetylcholine (ACh), acts on specific receptors on endothelial cells lining the luminal side of an arteriole. The receptor activation leads to the formation of compounds which cause relaxation of smooth muscle cells situated adjacent to the endothelial cells resulting in vasodilatation of the arteriole. B, detailed illustration of how vasodilator systems in the vascular wall can be activated and of proposed interactions between vasodilator systems. In vascular endothelial cells, several compounds, including acetylcholine, ATP, adenosine and bradykinin as well as mechanical signals including shear stress, activate both endothelial NO synthase (enos) and the arachidonic acid pathway leading to formation of prostacyclin (PGI 2 ) and eicosatrienoic acids (EETs). Activation of NOS and the arachidonic acid pathway occurs via an increase in intracellular calcium but enos activity is also regulated by protein phosphorylation at different sites. Redundancy exists between the NO and the prostacyclin systems where one described mechanism is inhibition of NOS by prostacyclin. Redundancy also exists between the NO system and cytochrome P450 2C9 (CYP 2C9), which produces 11,12-eicosatrienoic acid (11,12-EET) that can induce smooth muscle cell (SMC) relaxation by hyperpolarization. During normal conditions, NO exerts an inhibitory effect on CYP 2C9 but when NO formation is inhibited, the activity of CYP 2C9 increases. In the smooth muscle cell cyclic guanosine monophosphate (cgmp) can promote cyclic adenosine monophosphate (camp) levels by inhibiting phosphodiesterase III, which degrades camp. Abbreviations: AChR: acetylcholine receptor; AA: arachidonic acid; ATP: adenosine 5 -triphosphate; BKR: bradykinin receptor; 11, 12 EETs: 11, 12 eicosatrienoic acid; CaM: calmodulin; CaMK: calmodulin kinase; camp: cyclic adenosine monophosphate; cgmp: cyclic guanosine monophosphate; COX: cyclooxygenase; CYP 2C9: cytochrome P450 2C9; HR: histamine receptor; PGI 2 : prostacyclin; PKA, protein kinase A; PKC: protein kinase C; P1: purinergic receptor 1; P2: purinergic receptor 2; SMC: smooth muscle cell. Interactions of nitric oxide and prostanoids in endothelial cells In addition to their effects on smooth muscle cells, NO and prostanoids may also influence autacoid production in the endothelial layer. Inhibition of prostanoid synthesis leads to an increased NO formation in endothelial cells, an effect brought about by a camp induced reduction in the intracellular calcium level (Fig. 2) (Bolz & Pohl, 1997). The inhibitory effect of prostanoids on NO formation (Bolz & Pohl, 1997) could explain why exercise hyperaemia is not reduced when COX is inhibited in humans (Mortensen et al. 2007; 2009b). In this setting, inhibition of prostanoid formation would increase intracellular calcium, leading to a compensating increase in NO formation. In contrast, inhibition of the NO system may not severely affect NO function as this system is already suppressed by the contraction-induced formation of prostanoids (Frandsen et al. 2000; Karamouzis et al. 2001). This latter finding is supported by a lack of effect of NO synthase inhibition on the exercise induced increase in prostacyclin levels in the skeletal muscle interstitium (Frandsen et al. 2000). A redundancy interaction between the prostacyclin and the NO system may also explain the synergistic effect of combined NOS and COX inhibition in reducing ATP induced vasodilatation (Mortensen et al. 2009b). As described above, the NOS system also shows an interaction with CYP 2C9 in that inhibition of CYP 2C9 or NOS alone does not lower exercise hyperaemia, whereas the combined inhibition does (Hillig et al. 2003). The explanation may lie in the observation that NO can inhibit the activity of CYP 2C so that when NO is inhibited the activity of CYP2C is enhanced and more of the vasodilator 11,12-EET is formed (Bauersachs et al. 1996). Interactions of camp and cgmp in smooth muscle cells The intracellular second messengers involved in smooth muscle cell (SMC) relaxation are the cyclic nucleotides, cyclic adenosine monophosphate (camp) and cyclic C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

13 J Physiol Vasodilator interactions in skeletal muscle blood flow regulation 6301 guanosine monophosphate (cgmp), generated by the activity of adenylate and guanylate cyclase, respectively. cgmp is considered to be the main mediator of the cellular effectsproducedbyno(binaet al. 1994), although at least a part of the NO-induced SMC relaxation is cgmp independent (Carvajal et al. 2000), whereas the vasodilator effects of prostanoids are mediated via elevations in camp (Vanhoutte & Mombouli, 1996). Interestingly, cgmp has been shown to have an inhibitory effect on the degradation of camp (Maurice & Haslam, 1990), which may explain the synergistic effect of NO and prostanoids on SMC relaxation (Fig. 2) (de Wit et al. 1994). This synergism is physiologically interesting as it increases the sensitivity of the system, but on the other hand it also indicates that even a small attenuation in the formation of cgmp could have a large effect on the resulting dilatation. The latter suggestion could explain why the different vasodilator systems that stimulate cgmp and camp can be stimulated by so many different compounds. In this context, simultaneous inhibition of NO and prostanoid formation could depress SMC levels of camp and cgmp to such an extent that additional inhibition of vasodilator systems coupled to these cyclic nucleotides may not have afurthereffect. Vasodilator activation by ATP and adenosine Adenosine binds to P1 purinergic receptors whereas ATP binds to P2 receptors (Ralevic & Burnstock, 1998). The vasodilator effect of both adenosine (Ray et al. 2002; Mortensen et al. 2009b; Nyberg et al. 2010) and ATP (McCullough et al. 1997; Hammer et al. 2001; Mortensen et al. 2009a; Crecelius et al. 2011) have been shown to be mediated in part via formation of NO and prostanoids (Fig. 3). As ATP is degraded rapidly by membrane-bound and soluble nucleotidases within the vasculature (Gordon, 1986; Yegutkin, 2008), one explanation for the convergence of downstream signalling could be that the vasodilator effect of ATP is mediated via adenosine. However, inhibition of P1 receptors does not reduce the vasodilator response to intra-arterial ATP infusion in humans (Rongen et al. 1994; Mortensen et al. 2009b; Kirbyet al. 2010), suggesting that the vasodilator effect of intravascular ATP is independent of adenosine. This suggestion is also in congruence with observations in isolated endothelial cells demonstrating a similar potency of ATP and the P2Y receptor specific agonist UTP (da Silva et al. 2009; Raqeeb et al. 2011). Interestingly, in contrast to the adenosine-independent vasodilator effect of intravascular ATP, extraluminal application of ATP to blood-perfused arterioles has been suggested to be dependent on the action of adenosine on P1 receptors (Duza & Sarelius, 2003). This discrepancy between mechanisms underlying interstitial and intravascular ATP-induced vasodilatation is likely to reflect differences in receptor expression and/or the capacity for nucleotide degradation in the two compartments, but more evidence is needed to clarify the interaction of adenosine and ATP in the intravascular and interstitial space. Forearm and leg: do both models reflect skeletal muscle vasculature? Experimental considerations in the forearm versus leg model. Evaluation of vascular function and blood flow regulation relies on the determination of blood flow and/or changes in arterial vessel diameter which can be evaluated by various methods in the forearm and leg model (Casey et al. 2008). Because the forearm volume is low, smaller doses of first-pass pharmacological drugs can be infused into the arterial circulation, which reduces the risk of confounding systemic effects such as an increase in blood pressure. A drawback of the forearm model, however, is that arterio-venous differences across the forearm cannot be obtained, because there is no single vein draining the forearm (Wahren 1966). Consequently, release and uptake of substances across the experimental limb cannot be determined and forearm oxygen uptake ( V O2 ) cannot be estimated. Without forearm V O2,the physiological importance of differences in blood flow is undisclosed because it is not known if it is secondary to changes in vascular function or a change in metabolic demand. In regards to the latter, it is well known that some infused substances can alter local metabolism (Mortensen et al. 2007; Boushel et al. 2012). Limb specific vascular function. Due to the upright posture, the human legs are much more exposed to hydrostatic pressure when compared to the forearm (Rowell, 1993). In addition, the use of the leg muscles in locomotion holds the skeletal muscle tissue more active than the arm muscles, even in sedentary individuals. It therefore seems likely that there are limb differences in vascular function and blood flow regulation and studies that have compared the arm and leg have also reported differences in vasodilator responsiveness to endothelium-dependant and independent substances (Newcomer et al. 2004) and α 1 -adrenergic responsiveness (Pawelczyk & Levine, 2002). In the forearm, adenosine infusion shows a large difference in vasodilator effect among individuals (Martin et al. 2006) whereas the inter-subject variation is less in the leg (Mortensen et al. 2009b; Nyberg et al. 2010; Hellsten et al. 2012). Moreover, a large number of studies show that the vasodilator response to ACh is markedly reduced in the forearm in individuals with cardiovascular disease (Virdis et al. 2010); however, when measured in the leg, the vasodilator response to ACh has been found to be similar in hypertensive and normotensive individuals, and C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

14 6302 Y. Hellsten and others J Physiol in Type II diabetics and healthy controls (Thaning et al. 2011; Hellsten et al. 2012). This lack of difference in ACh induced vasodilatation also suggests that the finding that ACh infusion induces vasoconstriction in the forearm of individuals with cardiovascular disease, primarily by binding of prostacyclin to TP receptors (Félétou et al. 2010), is more evident in the arm. Ageing also appears to affect the endothelial function of the legs more than the arms (Thijssen et al. 2011). With regards to blood flow regulation, NO appears to be obligatory for exercise hyperaemia in the forearm of young subjects (Schrage, 2004), whereas it is not in the leg (Rådegran & Saltin, 1999; Frandsen et al. 2001; Nyberg et al. 2012). Notably, the capacity to form prostacyclin has also been shown to be lower in plasma and leg skeletal muscle in individuals with hypertension compared to normotensive control subjects (Hellsten et al. 2012). Thus, existing evidence suggest that there are differences in blood flow regulation in young healthy individuals and that ageing and cardiovascular diseases affect the arms and legs differently (Thijssen et al. 2011). Flow mediated dilatation in the forearm is widely used to evaluate endothelial function, but because of the apparent differences in vascular function and relative small mass of the forearm, it is likely that the leg or a combination between evaluation of the endothelial function of the arm and leg is a better indicator of the general cardiovascular system (Thijssen et al. 2011; Mortensen et al. 2012). More studies comparing Figure 3. Adenosine induced stimulation of NO and prostacyclin formation A, muscle interstitial nitrite and nitrate (NOx) and 6-PGF1α concentrations during baseline conditions and interstitial adenosine infusion through microdialysis probes. Significantly different from baseline (P < 0.05). B, effectof adenosine on release of NO from skeletal muscle and microvascular endothelial cells. Significant formation (P < 0.05). C, effect of adenosine on release of 6-PGF1α from skeletal muscle and microvascular endothelial cells. Significant formation (P < 0.05). Adapted from Nyberg et al. (2010). C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

15 J Physiol Vasodilator interactions in skeletal muscle blood flow regulation 6303 vascular function and regulation in the arms and legs and evaluating the prognostic value of the endothelial function in the arm and leg in regards to cardiovascular risks are needed, but existing data suggest that observations in one limb cannot be extrapolated to other limbs. Conclusion It is clear that regulation of skeletal muscle blood flow is a complex process involving several cellular sources, and many compounds and mechanisms. Most of the vasodilators proposed to be of importance in skeletal muscle blood flow regulation do not appear to be essential as their vasodilator effect can be compensated for by the formation of others. NO and prostacyclin are two central vasodilators which are stimulated by a number of other vasodilator compounds including ACh, ATP, adenosine and mechanical signalling. The many stimulators of NO and prostacyclin can probably be explained by redundancy and fine-tuned control of blood flow that ensures the vital delivery of O 2 even under compromised conditions. Pharmacological interventions in humans designed to reveal mechanisms underlying the regulation of skeletal muscle blood flow have succeeded in identifying specific vasodilators and vasodilator interactions of importance for exercise hyperaemia. However, regardless of the pharmacological interventions used, exercise hyperaemia has only been partially reduced, suggesting that yet unknown vasodilator mechanisms, potentially independent of camp and cgmp formation, remain to be revealed. Future studies should focus on the functional role of EDHF as well as potential novel mechanisms in muscle blood flow regulation. References Andersen P & Saltin B (1985). Maximal perfusion of skeletal-muscle in man. JPhysiol366, Baenziger NL, Force LE & Becherer PR (1980). Histamine stimulate prostacyclin synthesis in cultured human umbilical vein endothelial cells. Biochem Biophys Res Commun 92, Bauersachs J, Popp R, Hecker M, Sauer E, Fleming I & Busse R (1996). Nitric oxide attenuates the release of endothelium-derived hyperpolarizing factor. Circulation 94, BevegårdS&OröL (1969). Effect of prostaglandin E1 on forearm blood flow. Scand J Clin Lab Inevest 23, Bina S, Hart JL & Muldoon SM (1994). Comparative effects of exogenous nitrovasodilators on cgmp levels in different canine blood vessels. Life Sci 56, Bolz SS & Pohl U (1997). Indomethacin enhances endothelial NO release evidence for a role of PGI2 in the autocrine control of calcium-dependent autacoid production. Cardiovasc Res 36, Bolz SS, Fisslthaler B, Pieperhoff S, Wit CD, Fleming I, Busse R & Pohl U (2000). Antisense oligonucleotides against cytochrome P450 2C8 attenuate EDHF-mediated Ca 2+ changes and dilation in isolated resistance arteries. FASEB J 14, Boushel R, Fuentes T, Hellsten Y & Saltin B (2012). Opposing effects of nitric oxide and prostaglandin inhibition on muscle mitochondrial VO 2 during exercise. Am J Physiol Regul Integr Comp Physiol 303, R94 R100. Boushel R, Langberg H, Gemmer C, Olesen J, Crameri R, Scheede C, Sander M & Kjaer M (2002). Combined inhibition of nitric oxide and prostaglandins reduces human skeletal muscle blood flow during exercise. JPhysiol543, Bradley SJ, Kingwell BA & McConell GK (1999). Nitric oxide synthase inhibition reduces leg glucose uptake but not blood flow during dynamic exercise in humans. Diabetes 48, Busse R, Edwards G, Félétou M & Fleming I (2002). EDHF: bringing the concepts together. Trends Pharmacol Sci 23, Casey DP, Curry TB & Joyner MJ. Measuring muscle blood flow: a key link between systemic and regional metabolism. Curr Opin Clin Nutr Metab Care. (2008). 11(5): Carter TD, Hallam TJ, Cusack NJ & Pearson JD (1988). Regulation of P2y-purinoceptor-mediated prostacyclin release from human endothelial cells by cytoplasmic calcium concentration. Br J Pharmacol 95, Carvajal JA, Germain AM, Huidobro-Toro JP & Weiner CP (2000). Molecular mechanism of cgmp-mediated smooth muscle relaxation. JCellPhysiol184, Crecelius AR, Kirby BS, Voyles WF & Dinenno FA (2011). Augmented skeletal muscle hyperaemia during hypoxic exercise in humans is blunted by combined inhibition of nitric oxide and vasodilating prostaglandins. JPhysiol589, Clifford PS & Hellsten Y (2004). Vasodilatory mechanisms in contracting skeletal muscle. JApplPhysiol97, da Silva CG, Specht A, Wegiel B, Ferran C & Kaczmarek E (2009). Mechanism of purinergic activation of endothelial nitric oxide synthase in endothelial cells. Circulation 119, de Wit C, von Bismarck P & Pohl U (1994). Synergistic action of vasodilators that increase cgmp and camp in the hamster cremaster microcirculation. Cardiovasc Res 28, de Wit C, Schäfer C, von Bismarck P, Bolz SS & Pohl U (1997). Elevation of plasma viscosity induces sustained NO-mediated dilation in the hamster cremaster microcirculation in vivo. Pflugers Arch 434, Domeier TL & Segal SS (2007). Electromechanical and pharmacomechanical signalling pathways for conducted vasodilatation along endothelium of hamster feed arteries. J Physiol 579, Dudzinski D & Michel T (2007). Life history of enos: partners and pathways. Cardiovasc Res 75, Duza T & Sarelius IH (2003). Conducted dilations initiated by purines in arterioles are endothelium dependent and require endothelial Ca 2+. Am J Physiol Heart Circ Physiol 285, H26 H37. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

16 6304 Y. Hellsten and others J Physiol Dyke CK, Proctor DN, Dietz NM & Joyner MJ (1995). Role of nitric oxide in exercise hyperaemia during prolonged rhythmic handgripping in humans. JPhysiol488, Félétou M, Huang Y & Vanhoutte PM (2010). Vasoconstrictor prostanoids. Pflugers Arch 459, Félétou M, Huang Y & Vanhoutte PM (2011). Endothelium-mediated control of vascular tone: COX-1 and COX-2 products. Br J Pharmacol 164, Fisslthaler B, Popp R, Kiss L, Potente M, Harder DR, Fleming I & Busse R (1999). Cytochrome P450 2C is an EDHF synthase in coronary arteries. Nature 401, Fleming I & Busse R (2003). Molecular mechanisms involved in the regulation of the endothelial nitric oxide synthase. Am J Physiol Regul Integr Comp Physiol 284, R1 R12. Frandsen U, Bangsbo J, Langberg H, Saltin B & Hellsten Y (2000). Inhibition of nitric oxide synthesis by systemic N G -monomethyl- L-arginine administration in humans: effects on interstitial adenosine, prostacyclin and potassium concentrations in resting and contracting skeletal muscle. J Vasc Res 37, Frandsen U, Bangsbo J, Sander M, Hoffner L, Betak A, Saltin B & Hellsten Y (2001). Exercise-induced hyperaemia and leg oxygen uptake are not altered during effective inhibition of nitric oxide synthase with N G -nitro-l-arginine methyl ester in humans. JPhysiol531, Gilligan DM, Panza JA, Kilcoyne CM, Waclawiw MA, Casino PR & Quyyumi AA (1994). Contribution of endothelium-derived nitric oxide to exercise-induced vasodilation. Circulation 90, Gordon JL (1986). Extracellular ATP: effects, sources and fate. Biochem J 233, Grabowski EF, Jaffe EA & Weksler BB (1985). Prostacyclin production by cultured endothelial cell monolayers exposed to step increases in shear stress. JLabClinMed105, Halcox JP, Narayanan S, Cramer-Joyce L, Mincemoyer R & Quyyumi AA (2001). Characterization of endothelium-derived hyperpolarizing factor in the human forearm microcirculation. Am J Physiol Heart Circ Physiol 280, H2470 H2477. Hammer LW, Ligon AL & Hester RL (2001). ATP-mediated release of arachidonic acid metabolites from venular endothelium causes arteriolar dilation. Am J Physiol Heart Circ Physiol 280, H2616 H2622. Heinonen I, Bengt S, Jukka K, SipiläHT,VesaO,PirjoN, Juhani K, Kari K & Ylva H (2011). Skeletal muscle blood flow and oxygen uptake at rest and during exercise in humans: a pet study with nitric oxide and cyclooxygenase inhibition. Am J Physiol Heart Circ Physiol 300, Hellsten Y, Nyberg M & Mortensen SP (2012). Contribution of intravascular versus interstitial purines and nitric oxide in the regulation of exercise hyperaemia in humans. JPhysiol 590, Hillig T, Krustrup P, Fleming I, Osada T, Saltin B & Hellsten Y (2003). Cytochrome P450 2C9 plays an important role in the regulation of exercise-induced skeletal muscle blood flow and oxygen uptake in humans. JPhysiol(lond)546, Karamouzis M, Langberg H, Skovgaard D, Bülow J, Kjaer M & Saltin B (2001). In situ microdialysis of intramuscular prostaglandin and thromboxane in contracting skeletal muscle in humans. Acta Physiol Scand 171, Kingwell BA, Formosa M, Muhlmann M, Bradley SJ & McConell GK (2002). Nitric oxide synthase inhibition reduces glucose uptake during exercise in individuals with type 2 diabetes more than in control subjects. Diabetes 51, Kirby BS, Crecelius AR, Voyles WF & Dinenno FA (2010). Vasodilatory responsiveness to adenosine triphosphate in ageing humans. JPhysiol(Lond)588, Koller A, Sun D, Huang A & Kaley G (1994). Corelease of nitric oxide and prostaglandins mediates flow-dependent dilation of rat gracilis muscle arterioles. Am J Physiol Heart Circ Physiol 267, H326 H332. Martin EA, Nicholson WT, Eisenach JH, Charkoudian N & Joyner MJ (2006). Influences of adenosine receptor antagonism on vasodilator responses to adenosine and exercise in adenosine responders and nonresponders. JAppl Physiol 101, Maurice DH & Haslam RJ (1990). Molecular basis of the synergistic inhibition of platelet function by nitrovasodilators and activators of adenylate cyclase: inhibition of cyclic AMP breakdown by cyclic GMP. Mol Pharmacol 37, McCullough W, Collins D & Ellsworth M (1997). Arteriolar responses to extracellular ATP in striated muscle. Am J Physiol Heart Circ Physiol 272, H1886 H1891. Mortensen SP, Askew CD, Walker M, Nyberg M & Hellsten Y (2012). The hyperaemic response to passive leg movement is dependent on nitric oxide; a new tool to evaluate endothelial nitric oxide function. JPhysiol590, Mortensen SP, Gonzalez-Alonso J, Damsgaard R, Saltin B & Hellsten Y (2007). Inhibition of nitric oxide and prostaglandins, but not endothelial-derived hyperpolarizing factors, reduces blood flow and aerobic energy turnover in the exercising human leg. JPhysiol581, Mortensen SP, González-Alonso J, Bune LT, Saltin B, Pilegaard H & Hellsten Y (2009a). ATP-induced vasodilation and purinergic receptors in the human leg: roles of nitric oxide, prostaglandins, and adenosine. Am J Physiol Regul Integr Comp Physiol 296, R1140 R1148. Mortensen SP, Nyberg M, Thaning P, Saltin B & Hellsten Y (2009b). Adenosine contributes to blood flow regulation in the exercising human leg by increasing prostaglandin and nitric oxide formation. Hypertension 53, Mount PF, Kemp BE & Power DA (2007). Regulation of endothelial and myocardial NO synthesis by multi-site enos phosphorylation. JMolCellCardiol42, Nathan C & Xie QW (1994). Nitric oxide synthases: roles, tolls, and controls. Cell 78, Newcomer SC, Leuenberger UA, Hogeman CS, Handly BD & Proctor DN (2004). Different vasodilator responses of human arms and legs. JPhysiol556, Nyberg M, Jensen LG, Thaning P, Hellsten Y & Mortensen SP (2012). Role of nitric oxide and prostanoids in the regulation of leg blood flow and blood pressure in humans with essential hypertension: effect of high-intensity aerobic training. JPhysiol590, C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

17 J Physiol Vasodilator interactions in skeletal muscle blood flow regulation 6305 Nyberg M, Mortensen SP, Thaning P, Saltin B & Hellsten Y (2010). Interstitial and plasma adenosine stimulate nitric oxide and prostacyclin formation in human skeletal muscle. Hypertension 56, Panza JA, Casino PR, Badar DM & Quyyumi AA (1993). Effect of increased availability of endothelium-derived nitric oxide precursor on endothelium-dependent vascular relaxation in normal subjects and in patients with essential hypertension. Circulation 87, Pawelczyk JA & Levine BD (2002). Heterogeneous responses of human limbs to infused adrenergic agonists: a gravitational effect? JApplPhysiol92, Ralevic V, Burnstock G. (1998). Receptors for purines and pyrimidines.pharmacol Rev. 50(3): Review. Raqeeb A, Sheng J, Ao N & Braun AP (2011). Purinergic P2Y2 receptors mediate rapid Ca 2+ mobilization, membrane hyperpolarization and nitric oxide production in human vascular endothelial cells. Cell Calcium 49, Ray C & Marshall J (2006). The cellular mechanisms by which adenosine evokes release of nitric oxide from rat aortic endothelium. JPhysiol570, Ray CJ, Abbas MR, Coney AM & Marshall JM (2002). Interactions of adenosine, prostaglandins and nitric oxide in hypoxia-induced vasodilatation: in vivo and in vitro studies. J Physiol 544, Rådegran G & Saltin B (1999). Nitric oxide in the regulation of vasomotor tone in human skeletal muscle. Am J Physiol Heart Circ Physiol 276, H1951 H1960. Rongen GA, Smits P & Thien T (1994). Characterization of ATP-induced vasodilation in the human forearm vascular bed. Circulation 90, Rowell LB (1993). Human Cardiovascular Control.Oxford University Press, New York. Sarelius I & Pohl U (2010). Control of muscle blood flow during exercise: local factors and integrative mechanisms. Acta Physiologica 199, Schrage WG (2004). Local inhibition of nitric oxide and prostaglandins independently reduces forearm exercise hyperaemia in humans. JPhysiol557, Schrage WG, Wilkins BW, Johnson CP, Eisenach JH, Limberg JK, Dietz NM, Curry TB & Joyner MJ (2010). Roles of nitric oxide synthase and cyclooxygenase in leg vasodilation and oxygen consumption during prolonged low-intensity exercise in untrained humans. JApplPhysiol109, Shoemaker JK, Naylor HL, Pozeg ZI & Hughson RL (1996). Failureofprostaglandinstomodulatethetimecourseof blood flow during dynamic forearm exercise in humans. J Appl Physiol 81, Thaning P, Bune LT, Zaar M, Saltin B & Rosenmeier JB (2011). Functional sympatholysis during exercise in patients with type2diabeteswithintactresponsetoacetylcholine. Diabetes Care 34, Thijssen DHJ, Rowley N, Padilla J, Simmons GH, Laughlin MH, Whyte G, Cable NT & Green DJ (2011). Relationship between upper and lower limb conduit artery vasodilator function in humans. JApplPhysiol111, Vallance P, Collier J & Moncada S (1989). Effects of endothelium-derived nitric oxide on peripheral arteriolar tone in man. Lancet 334, Vanhoutte PM & Mombouli JV (1996). Vascular endothelium: vasoactive mediators. Prog Cardiovasc Dis 39, Vanhoutte PM, Shimokawa H, Tang EHC & Félétou M (2009). Endothelial dysfunction and vascular disease. Acta Physiologica 196, Virdis A, Ghiadoni L & Taddei S (2010). Human endothelial dysfunction: EDCFs. Pflugers Arch 459, Wahren J (1966). Quantitative aspects of blood flow and oxygen uptake in the human forearm during rhythmic exercise. Acta Physiol Scand Suppl 269, Wilson JR & Kapoor SC (1993). Contribution of prostaglandins to exercise-induced vasodilation in humans. Am J Physiol Heart Circ Physiol 265, H171 H175. Yegutkin GG (2008). Nucleotide- and nucleoside-converting ectoenzymes: Important modulators of purinergic signalling cascade. Biochim Biophys Acta 1783, Acknowledgements Funding for the referenced work of the authors was gratefully received from The Danish Council for Independent Research Medical Sciences, Lundbeck Foundation, Novo Nordisk Foundation, The Danish Ministry of Culture, The Danish Heart Association and P. Carl Petersen Foundation. C 2012 The Authors. The Journal of Physiology C 2012 The Physiological Society Downloaded from J Physiol (jp.physoc.org) at Copenhagen University Library on December 28, 2012

18 Autonomic Adjustments to Exercise in Humans James P. Fisher, 1 Colin N. Young, 2 and Paul J. Fadel 3,4* ABSTRACT Autonomic nervous system adjustments to the heart and blood vessels are necessary for mediating the cardiovascular responses required to meet the metabolic demands of working skeletal muscle during exercise. These demands are met by precise exercise intensity-dependent alterations in sympathetic and parasympathetic nerve activity. The purpose of this review is to examine the contributions of the sympathetic and parasympathetic nervous systems in mediating specific cardiovascular and hemodynamic responses to exercise. These changes in autonomic outflow are regulated by several neural mechanisms working in concert, including central command (a feed forward mechanism originating from higher brain centers), the exercise pressor reflex (a feed-back mechanism originating from skeletal muscle), the arterial baroreflex (a negative feed-back mechanism originating from the carotid sinus and aortic arch), and cardiopulmonary baroreceptors (a feed-back mechanism from stretch receptors located in the heart and lungs). In addition, arterial chemoreceptors and phrenic afferents from respiratory muscles (i.e., respiratory metaboreflex) are also capable of modulating the autonomic responses to exercise. Our goal is to provide a detailed review of the parasympathetic and sympathetic changes that occur with exercise distinguishing between the onset of exercise and steady-state conditions, when appropriate. In addition, studies demonstrating the contributions of each of the aforementioned neural mechanisms to the autonomic changes and ensuing cardiac and/or vascular responses will be covered. C 2015 American Physiological Society. Compr Physiol 5: , Introduction The autonomic nervous system plays a critical role in mediating the cardiovascular adjustments necessary to meet the metabolic demands of the exercising muscle, and as such is paramount for the performance and sustainment of physical activity. A reduction in the tonic suppressive influence of parasympathetic (vagus) nerve activity contributes to exercise-induced increases in heart rate (HR), ventricular contractility, stroke volume, and thus, cardiac output. Increases in HR and ventricular contractility are also evoked by activation of cardiac sympathetic nerve activity (SNA) and sympathetic stimulation of epinephrine release from the adrenal medulla. In addition, a sympathetically mediated vasoconstriction in nonexercising muscles and visceral organs (e.g., splanchnic circulation) facilitates the redistribution of cardiac output to the active skeletal muscles. At the same time, the normal ability of SNA to cause vasoconstriction is attenuated in the active muscles, in part due to an effect of muscle metabolites to diminish the vasoconstrictor response to α-adrenergic receptor activation (123,268). Such modulation, termed functional sympatholysis, may constitute a protective mechanism that optimizes muscle blood flow in the face of the increased sympathetic vasoconstrictor drive that occurs during exercise. However, sympatholysis is not complete and the increased SNA to active muscles does have a restrictive effect on blood flow, which is important for the maintenance of arterial blood pressure (BP) during exercise (337). Overall, the importance of the autonomic adjustments to exercise can be readily appreciated from the finding that patients with autonomic failure cannot maintain the lightest loads of dynamic exercise even if performed in the supine position to enhance central blood volume and venous return (199). The autonomic responses to exercise are dependent upon the type of exercise that is being performed. In general, exercise can be divided into two categories: dynamic and isometric (or static) (185, 221). Dynamic exercise involves rhythmical contractions that alter both muscle length and joint angle and involves intermittent changes in intramuscular force that occur in conjunction with the contraction and relaxation of the working muscles. This intermittent pumping action of skeletal muscle (i.e., muscle pump) contributes to increases in muscle blood flow. Isometric exercise on the other hand involves a sustained contraction with minimal change in muscle length or joint angle and includes a substantial development of intramuscular force. The large increases in intramuscular pressure are transferred to the vasculature and cause a decrease in * Correspondence to fadelp@health.missouri.edu 1 School of Sport, Exercise & Rehabilitation Sciences, College of Life & Environmental Sciences, University of Birmingham, Birmingham, United Kingdom 2 Biomedical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY, USA 3 Department of Medical Pharmacology & Physiology, University of Missouri, Columbia, MO, USA 4 Dalton Cardiovascular Research Center, University of Missouri, Columbia, MO, USA Published online, April 2015 (comprehensivephysiology.com) DOI: /cphy.c Copyright C American Physiological Society. Volume 5, April

19 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology skeletal muscle blood flow. These contrasting contractile characteristics and the subsequent hemodynamic alterations contribute importantly to the differences in the cardiovascular responses evoked. While HR and BP increase with both dynamic and isometric exercise, there are distinct differences. At a constant work load, during dynamic exercise HR increases to a steady-state value, whereas during isometric exercise HR continually rises at a given work load until fatigue. However, perhaps the most notable difference between dynamic and isometric exercise is the pressor response, which occurs to a much greater extent during isometric exercise owing to the more immediate and large increases in SNA with this form of exercise, as discussed in detail later in this review. Studies designed to examine the autonomic responses to exercise have incorporated both dynamic and isometric exercise protocols. Handgrip and knee extensions performed as a percent of maximal voluntary contraction (% MVC) are the primary forms of isometric exercise used. The classic mode of dynamic exercise for investigation is cycling; however, laboratory studies investigating neural cardiovascular control during dynamic exercise more often use a rhythmical form of isometric exercise which incorporates intermittent isometric contractions separated by timed relaxation periods. A major advantage of this form of exercise is that it more readily allows for movement sensitive measurements such as muscle SNA (microneurography) or beat-to-beat blood flow (Doppler ultrasound) to be obtained without the motion artifacts inherent to whole body dynamic exercise. Likewise, the use of dynamic and isometric exercise in human research protocols has been instrumental in teasing apart the contributions of the underlying neural mechanisms involved in evoking the autonomic and ensuing cardiovascular responses to exercise. Several neural mechanisms working in concert are responsible for the autonomic adjustments to exercise and through complex interactions precisely control the cardiovascular and hemodynamic changes in an intensity-dependent manner (Fig. 1). It is well accepted that central signals from the higher brain associated with the volitional component of exercise (i.e., central command) (264,362,378), peripheral signals arising from mechanically and metabolically sensitive afferents in contracting skeletal muscle (i.e., exercise pressor reflex) (162, 164, 219, 220, 315), and feedback from stretch receptors originating in the carotid and aortic arteries (i.e., arterial baroreflex) (75, 76, 78, 263, 265) are all involved. Less appreciated, but also important, are low-pressure mechanically sensitive stretch receptors located in the heart, great veins and blood vessels of the lungs that sense changes in central blood volume and pressure (i.e., the cardiopulmonary baroreflex) (41, 76, 78, 88, 163, 209). In addition, arterial chemoreceptors housed in the carotid and aortic bodies and phrenic afferents from respiratory muscles (i.e., respiratory metaboreflex) are also capable of modulating the autonomic responses to exercise. This review will focus on the contributions of the Central command Carotid chemoreflex Medullary cardiovascular areas PSNA Ach Arterial baroreceptors SNA SNA SNA Norepi Ach Heart Exercise pressor reflex Cardiopulmonary baroreceptors Respiratory metaboreflex Norepi Peripheral vasculature Adrenal medulla Figure 1 Schematic summarizing the mechanisms involved in mediating the autonomic adjustments to exercise. Neural signals originating from higher brain centers (i.e., central command), chemically sensitive receptors in the carotid and aortic bodies (i.e., arterial chemoreflex), stretch receptors in the carotid and aortic arteries (i.e., arterial baroreflex), mechanically and metabolically sensitive afferents from skeletal muscle (i.e., exercise pressor reflex), mechanically sensitive stretch receptors in the cardiopulmonary region (i.e., cardiopulmonary baroreflex) and metabolically sensitive afferents from respiratory muscles (i.e., respiratory metaboreflex) are processed within brain cardiovascular control areas that influence efferent sympathetic and parasympathetic nerve activity. The alterations in autonomic outflow elicited by these inputs during exercise evoke changes in cardiac and vascular function, as well as release of catecholamines from the adrenal medulla. 476 Volume 5, April 2015

20 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans sympathetic and parasympathetic nervous systems in mediating specific cardiovascular and hemodynamic responses to exercise. Studies demonstrating the contributions of the aforementioned neural mechanisms to the autonomic changes and ensuing cardiac and/or vascular responses will be covered. In addition, distinctions between the onset of exercise and steady-state conditions will be made, when appropriate. Of note, nonadrenergic hormonal contributions (e.g., Angiotensin II, Vasopressin) to the cardiovascular responses to exercise will not be covered nor will autonomic effects on metabolic and respiratory responses. Overall, the literature examining the autonomic nervous system and the mechanisms of neural cardiovascular control during exercise is vast. As such, it is not feasible to include all of the research within these fields in a single article and therefore, additional review articles are cited throughout to direct the reader to further work in these important areas. It should also be noted that this review was designed to focus primarily on studies in healthy humans and more recently published work with only brief mention to animal studies and historical data. Autonomic Responses to Exercise Cardiac autonomic regulation Following a brief description of the organization of the autonomic neural control of the heart and the methods commonly used for its assessment in humans, the influence of exercise phase (e.g., onset, steady-state, and recovery), duration, intensity, and modality on cardiac autonomic control will be discussed. Cardiac autonomic organization Parasympathetic and sympathetic efferent activity can modulate the chronotropic, inotropic, and lusitropic functioning of the heart. The cell bodies of the parasympathetic preganglionic neurons are situated in the nucleus ambiguus and dorsal motor nucleus of the medulla oblongata (154). The axons travel within the tenth cranial nerve (vagus nerve) and synapse at a postganglionic neuron located at the cardiac plexus. Sympathetic preganglionic cell bodies are located in the intermediolateral cell column of the spinal cord and the preganglionic fibers directed to the heart synapse at the stellate ganglion and upper thoracic ganglia (T1-T5). Postganglionic parasympathetic and sympathetic fibers synapse at the sinoatrial node, atrioventricular node, atria, and ventricles. The classical view of cardiac autonomic neurotransmission whereby parasympathetic efferents release acetylcholine onto muscarinic receptors and sympathetic efferents release norepinephrine onto β 1 -receptors is well established (154). However, the importance of the complex pre- and postsynaptic interactions between parasympathetic and sympathetic fibers (e.g., excitatory facilitation and accentuated antagonism) (189) and the role of intrinsic and locally released neuromodulators (e.g., neuronal nitric oxide and neuropeptide Y) (18, 136) should not be overlooked. Although the rich sympathetic innervation of ventricles is widely recognized, as recently reviewed by Coote (39), there is accumulating evidence for dense parasympathetic innervation of this region and its functional significance in the control of ventricular rhythm, rate and contractility. In addition, a positive chronotropic response to α 1 -adrenergic stimulation has been reported in young individuals (287). However, the importance of parasympathetic regulation of ventricular function and α 1 - adrenoreceptor chronotropism during exercise in health and disease remains to be elucidated. Assessment of cardiac autonomic regulation The limited accessibility of cardiac autonomic efferents means that obtaining direct intraneural recordings from humans is not plausible, thus indirect assessments of cardiac autonomic control must be relied upon. The relative merits and faults of these experimental approaches have been discussed extensively elsewhere (32) and will only be briefly mentioned here. The use of pharmacological blocking agents provides a valuable means of dissecting the contributions of parasympathetic and sympathetic activity to cardiac regulation. Cardiac parasympathetic activity may be assessed by the administration of muscarinic receptor antagonists such as atropine sulfate or glycopyrronium bromide (also known as glycopyrrolate), with the latter being preferred in more recent investigations as its penetration of the blood-brain barrier is much lower, which decreases the likelihood of any potential confounding central effects (259). The pharmacological assessment of cardiac sympathetic activity may be achieved by the administration of β-adrenergic blockade. Although β- blockade has been usually administered prior to exercise, a small number of studies in animals have initiated administration during exercise (22, 239), thus circumventing any confounding effects of a change in baseline HR. The potentially confounding effects of drug non-specificity (e.g., combined β 1 - and β 2 -adrenergic receptor blockers) and blockade completeness should be considered when interpreting pharmacological investigations of cardiac autonomic control during exercise. Also, while very useful in the assessment of HR regulation during exercise, the associated changes in diastolic filling time and thus preload, means that pharmacological examination of the inotropic effects of cardiac autonomic activity in vivo can be difficult. On the basis that cardiac SNA is proportional to the appearance of norepinephrine in the coronary venous effluent (coronary sinus); cardiac norepinephrine spillover rate can provide an accurate assessment of cardiac sympathetic firing (169). Norepinephrine spillover from the heart may be studied with the infusion of radio-labeled norepinephrine and appropriate catheterization and regional blood sampling (74). To the best of the authors knowledge, a comparable approach has not been used to assess cardiac acetylcholine spillover from the human heart, which would require appropriate control for the high catalytic activity Volume 5, April

21 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology of acetylcholinesterase (309). While the directness of this approach is a key strength, its invasiveness and the associated technical challenges limits it widespread use. Albeit much more indirect, time and frequency domain analysis of HR variability provides a noninvasive means of assessing cardiac autonomic activity (69). As power spectral analysis of fluctuations in R-R interval occurring around a respiratory frequency (i.e., high frequency, Hz) and short-term time domain measures (e.g., root mean square of successive differences) of HR variability are virtually abolished by cholinergic-muscarinic blockade (69, 262) they are widely utilized to estimate cardiac parasympathetic activity. In contrast, low frequency (i.e., Hz) R-R interval power is attenuated by both cholinergic and β-adrenergic blockade suggesting that the activities of the parasympathetic and sympathetic nerves contribute, and thus the interpretation of changes in low frequency power is complex (69,169). The specific limitations associated with the use of HR variability to interrogate cardiac autonomic control during exercise have been discussed at length elsewhere (32). Patients who possess a functionally denervated heart as a consequence of spinal cord transection or heart transplantation also provide a valuable means of interrogating cardiac autonomic control in humans. Cardiac adrenergic innervation and activation can also be assessed with radio-scanning methods (74). This approach has been used to document heightened cardiac sympathetic activity in chronic heart failure patients and is also associated with elevated SNA directed to the skeletal muscle vasculature, reduced exercise capacity and future cardiac events in this population (387, 388). Although providing a powerful research tool, limited studies have used this approach to evaluate cardiac SNA during acute exercise (10), possibly due to its expense and poor temporal resolution. Cardiac autonomic regulation at exercise onset HR increases virtually instantaneously upon the initiation of exercise. The involvement of a reduction in cardiac parasympathetic activity (or inhibition of the normal restraining action of the inhibitory centre ) in humans was suggested at the turn of the last century (26) although experimental support for this proposition is much more recent (343). Administration of atropine has been shown to significantly attenuate the initial increase in HR to a variety of exercise modalities, including isometric handgrip (98, 194), isometric arm flexion (143), and leg cycling (80,271). Furthermore, HR variability derived indices of cardiac parasympathetic activity are decreased in early exercise (11, 22). In contrast, the magnitude of the cardiac acceleration at the onset of leg cycling is not diminished by prior β-adrenergic blockade, implying a minimal influence of sympathetic activity at this time (80, 271). Taken together these findings suggest that the early HR response to exercise is principally mediated by the withdrawal of cardiac parasympathetic activity, with the sympathetic contribution being manifest at a longer latency. This concept is supported by investigations showing a rapid reduction in HR resulting from stimulation of cardiac parasympathetic efferents and a more sluggish response to sympathetic nerve stimulation ( 3-6 s) (133). Contrary to the traditional view, Matsukawa and colleagues have advanced the hypothesis that it is an increase in cardiac SNA at the onset of exercise that principally causes the initial rise in HR. This view is supported by studies in conscious cats in which directly recorded cardiac SNA was shown to rapidly increase at the onset of treadmill exercise (by 168%-297% within 7 s) (346). However, reconciling these observations in cats with the human studies described above is challenging in the absence of direct cardiac sympathetic recordings in humans. Furthermore, alternative techniques for directly assessing cardiac SNA in humans do not offer sufficient temporal resolution to capture the kinetics of the onset response (e.g., cardiac plasma norepinephrine spillover). It has also been shown that tetraplegic subjects, in whom a complete cervical spinal cord lesion (C6-C7) has caused cardiac sympathetic denervation while cardiac parasympathetic innervation remains intact, display an attenuated increase in HR at the start of isometric arm exercise (329). Furthermore, the normal reduction in R-R interval high frequency spectral power is more sluggish in these patients during isometric exercise, compared to control participants (328). These observations have been taken as further evidence for the relative importance of SNA to the initial cardiac acceleration that accompanies exercise. However, an alteration in cardiac parasympathetic regulation as a consequence of a chronic adaptation to spinal lesion cannot be excluded. Cardiac autonomic regulation during steady-state exercise There is an approximately linear relationship between HR and oxygen uptake during incremental dynamic exercise of a large muscle mass (180). It is generally considered that the relative contribution of cardiac parasympathetic withdrawal to this HR response is greatest at lower exercise workloads and becomes less significant as exercise intensity increases, particularly once HR exceeds 100 b min 1. Conversely, the relative contribution of cardiac sympathetic stimulation to the exercise-induced elevation in HR increases along with exercise workload. Indeed, administration of atropine diminishes the magnitude of the HR rise during low-to-moderate intensity exercise, whereas HR is unaffected by propranolol administration (80,194,271). In contrast, at higher exercise intensities β- adrenergic blockade significantly attenuates the size of the HR response, thus suggesting a key role for heightened SNA and circulating catecholamines at such workloads (80, 194, 271). An exercise intensity-dependent reduction in high frequency power spectral density has also been described, indicative of progressive cardiac parasympathetic withdrawal (11). This is much more marked in the transition from rest to low-intensity exercise, than between moderate to high exercise intensity in healthy individuals (11). Such changes in autonomic modulation of the heart were absent in heart transplant patients 478 Volume 5, April 2015

22 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans possessing a denervated heart (11). Nevertheless, the relative contributions of cardiac parasympathetic activity and SNA to the HR response at differing dynamic exercise intensities continues to be debated (374). HR variability derived indices of cardiac parasympathetic activity have also been demonstrated to decrease during isometric exercise (149), although the magnitude of this reduction is less marked than observed during dynamic exercise and typically accompanied by a more modest increase in HR (128). As mentioned above, HR variability analyses do not permit a robust assessment of cardiac SNA, and to the authors knowledge, the influence of dynamic exercise intensity on cardiac norepinephrine spillover has not been comprehensively evaluated in healthy humans. However, Hasking et al. (129) has reported that cardiac norepinephrine spillover was markedly increased during moderate intensity leg cycling exercise (from 5 ± 2to73± 23 ng min 1 ). During prolonged submaximal dynamic exercise at a steady-state workload, there is a progressive increase in HR, mirrored by a fall in stroke volume, such that cardiac output is maintained relatively stable (43, 66). This cardiovascular drift phenomenon is exacerbated when exercise is combined with heat stress and dehydration (111). Intriguingly, β 1 -adrengeric blockade prevented the normal increase in HR observed after 15 min of leg cycling exercise ( 57% peak oxygen uptake, thermo-neutral conditions) and consequently stroke volume did not fall, likely as a consequence of the relatively longer diastolic filling time (100). Whether cardiac norepinephrine spillover increases over the time course of such exercise is unclear, however in dogs performing prolonged steady-state submaximal treadmill exercise the progressive increase in HR was accompanied by a progressive reduction in HR variability determinants of cardiac parasympathetic activity (182). An increase in core temperature may explain the autonomic contribution to the drift in HR noted during prolonged submaximal exercise (100, 157), although increased central command and skeletal muscle afferent feedback likely also contribute (178). The heart appears to retain at least some parasympathetic control at high exercise intensities. Indeed, atropine administration elicits a robust increase in HR in dogs performing heavy-intensity treadmill exercise (239) and respiratory mediated fluctuations in HR are still evident (22). However, the lack of an increase in HR with atropine administration during exhaustive dynamic exercise and the absence of appreciable HR variability in humans, suggests that cardiac parasympathetic withdrawal is complete at these workloads (272). An early study by Robinson et al. (272) reported that cardiac parasympathetic blockade reduced maximal oxygen uptake, although this observation was not substantiated by later investigations (65). As recently reviewed, an extensive body of work has evaluated the contribution of cardiac SNA to maximal exercise performance using a variety of human and animal models (343). The most relevant studies with respect to the focus of the present review are those employing β-adrenergic blockade in humans. Several of these studies have reported a reduction in maximal oxygen consumption and exercise capacity (9, 72, 336); however, this has not been a universal finding with no effect of β-adrenergic blockade often reported (65, 196, 269). Part of the reason for these equivocal findings may relate to the training status of the participants studied. Joyner et al. (159) demonstrated that β-adrenergic blockade with propranolol had little effect on maximal oxygen uptake in untrained subjects ( 45 ml min 1.kg 1 ) whereas in aerobically trained individuals ( 63 ml min 1.kg 1 ) a notable reduction in maximal oxygen uptake was observed. While, the importance of a direct cardiac effect is clear from the reported reduction in maximal cardiac output, indirect mechanisms may also contribute (e.g., metabolic and hormonal alterations) particularly when nonspecific β-adrenergic antagonists have been utilized. Along with a reduction in intrinsic HR, a reduction in β-adrenergic sensitivity contributes to the well-established reduction in maximal HR with age (33) that makes a significant contribution to the age-related lowering of maximal oxygen consumption (134). Maximal HR is often estimated as 220 age (93), but may underestimate maximal HR in older individuals and on the basis of extensive meta-analyses and laboratory-based investigations alternative regression equations have been derived (e.g., Tanaka equation; age) (270, 333). An inability of HR to increase appropriately in proportion to the metabolic demands of exercise has been termed chronotropic incompetence and has been linked with an impairment in β-adrenergic sensitivity (37, 71). It may be defined in a number of ways, but the failure to attain 80% of HR reserve during an incremental maximal exercise test is most commonly used (28). Importantly, chronotropic incompetence is associated with exercise intolerance and is an independent predictor of mortality in clinical (e.g., heart failure) and healthy populations (28, 71). It should be noted that while alterations in cardiac autonomic control may be implicated in chronotropic incompetence the underlying mechanisms remain incompletely understood. At the immediate cessation of exercise the initial recovery of HR is quite abrupt, but this is generally followed by a more gradual decline occurring over minutes, the specific kinetics of which are dependent on the intensity and duration of the prior exercise (130, 150). HR recovery is delayed in heart transplant patients (292) indicative of a contribution of cardiac autonomic efferent activity in this process. The rapid recovery of HR following exercise has been attributed to the rapid restoration of cardiac parasympathetic activity, since it is virtually abolished by atropine administration, but unaffected by β-adrenergic blockade (87, 150). Furthermore, endurance trained athletes in whom cardiac parasympathetic activity is elevated demonstrate a more rapid recovery of HR (150). The secondary more gradual decline in HR following exercise may be attributable to a slower restoration of the remaining cardiac parasympathetic activity and reduction in cardiac sympathoexcitation (11), which sustains a modest elevation in cardiac output, thus preserving perfusion pressure in the face of peripheral vasodilatation. HR recovery kinetics have been shown to have prognostic value, with a poor Volume 5, April

23 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology recovery (e.g., a fall in peak exercise HR by <12 b.min 1 in 1 min of supine recovery) being a predictor of an increase in all-cause mortality (34, 360). The reason for this perhaps relates to the association with the cardioprotective effects of cardiac parasympathetic activity (347). In summary, the available evidence generally supports the view that increases in HR during low intensity steady-state dynamic exercise are principally driven by a reduction in cardiac parasympathetic activity, whereas the relative contribution of a sympathetic mechanism is greater as exercise intensity increases. However, this viewpoint should not be considered absolute as some parasympathetic control of HR has been reported during higher intensity exercise. During prolonged dynamic exercise at a steady-state submaximal workload a sympathetically mediated HR rise is evoked (cardiac drift), which is exacerbated by dehydration and high ambient temperature. Furthermore, a reduction in β-adrenergic receptor sensitivity has been associated with a failure to normally increase HR and a reduction in aerobic exercise capacity. Thus, a normally functioning autonomic nervous system is requisite for an appropriate cardiac response during exercise and has important implications for exercise performance. Sympathetic regulation of the periphery Although both parasympathetic and sympathetic activity contribute to the cardiac responses to exercise, it is the sympathetic nervous system that is essential for the peripheral vascular adjustments to exercise. Here, we will provide a brief description of the organization of the sympathetic innervation of the peripheral vasculature and methods used for its assessment in humans. This will be followed by discussion of the current understanding of the sympathetic adjustments to exercise onset and steady-state conditions of varying intensity and duration. Peripheral sympathetic organization The tonic rhythmic discharge of sympathetic nerves is a major contributor to resting vasomotor tone in the peripheral circulation, and via modulation of the arterial baroreflex plays an essential role in BP homeostasis (42). However, the sympathetic nervous system is not only important for vascular tone and BP control under resting conditions, but is also intimately involved in the regulation of these processes during exercise. The central regulation of sympathetic outflow occurs mainly within cardiovascular areas of the brainstem (i.e., medulla oblongata) (50) (Fig. 2). Sympathetic preganglionic neurons are located in the intermediolateral cell column (IML) of the thoracic and upper two lumbar segments of the spinal cord (T1 - L2; i.e., thoracolumbar). These preganglionic neurons are cholinergic, fast conducting ( 15 m/s), thinly myelinated fibers that project via the ventral roots and the white rami to then synapse on postganglionic neurons in the paravertebral and prevertebral ganglia. The postganglionic neurons are noradrenergic, slower conducting ( 1 m/s), unmyelinated Circulating factors Arterial baroreceptors Skeletal muscle afferents Cardiopulmonary baroreceptors Respiratory muscle afferents Carotid chemoreceptors Medulla oblongata Circumventricular orgens NTS NA + CVLM Hypothalamic nuclei Periaqueductal gray Pontine structures Amygdala Cortical structures RVLM Hypothalamic nuclei Pontine structures Ventromedial medulla Caudal raphe nuclei Spinal interneurones Parasympathetic outflow IML Sympathetic outflow Figure 2 Simplified schematic illustrating putative neural areas involved in the control of parasympathetic and sympathetic outflow. The activity of efferent autonomic outflow is determined by complex interactions within and between central neural circuits, peripheral afferent inputs to the nucleus tractus solitarii (NTS) and other neural areas (not shown) as well as circulating factors. The integrated response of all of these ascending and descending neural signals ultimately determines parasympathetic and sympathetic outflow. See text for additional details. CVLM, caudal ventrolateral medulla; CVO, circumventricular organs; IML, intermediolateral cell column; NA, nucleus ambiguus; RVLM, rostral ventrolateral medulla. 480 Volume 5, April 2015

24 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans fibers that project through gray rami, and peripheral nerves to innervate target organs such as the heart and peripheral blood vessels. The sympathetic preganglionic neurons in the IML receive strong excitatory drive from neurons of the rostral ventrolateral medulla (RVLM) (50,324). Sympathetic preganglionic neurons at the IML also receive direct excitatory inputs from other regions of the central nervous system including the ventromedial medulla, caudal raphe nuclei, A5 noradrenergic cell group of the caudal ventrolateral pons, and the paraventricular hypothalamic nucleus (50,324). The regulation of this excitatory drive to the IML is multifaceted and can be intrinsically generated, hormonally mediated (e.g., Angiotensin II), or attributable to integration of excitatory inputs from other regions of the central nervous system (50). The circumventricular organs, which lack a blood brain barrier, provide an additional means by which circulating factors can influence key central regions involved in the regulation of SNA (310). While activation of sympathetic preganglionic neurons would lead to greater sympathetic outflow and norepinephrine release in the periphery, another important consequence of activating the sympathetic nervous system, particularly with exercise, is the stimulation of the release of epinephrine from the adrenal medulla and its impact on the cardiac and peripheral adjustments to exercise. However, our focus in this review will be on postganglionic sympathetic fibers to target organs and norepinephrine release with adrenal medulla stimulation and epinephrine effects only being discussed in brevity. Several excellent original articles and reviews are available for additional detail and description of this research literature (27, 103, 170, 171, 207, 389). In any description of the control of the sympathetic nervous system, the arterial baroreflex warrants discussion because of its profound beat-to-beat influence on SNA (138, 158, 197, 340). Arterial baroreceptor afferents emanating from mechanoreceptors at the carotid sinus bifurcation and the aortic arch terminate at the nucleus tractus solitarii (NTS) in the medulla oblongata, a major center for autonomic integration in the brainstem that receives both ascending and descending projections. The NTS exerts a tonic sympathoinhibitory influence on RVLM neurons by its projections to the caudal ventrolateral medulla (50). It is also important to consider that afferent inputs from other peripheral reflexes such as the arterial chemoreflex can impinge on the NTS to modulate sympathetic outflow (295) (Fig. 2). An alteration in the activity of any of these neural areas, and/or the reciprocal interactions amongst them, has the potential to modulate central sympathetic outflow at rest as well as during exercise. Indeed, the complex neuroanatomical interactions that contribute to the central regulation of SNA continue to be an intense area of investigation. Peripheral sympathetic assessment There are a multitude of methodologies for quantification of the activity of the sympathetic nervous system (74, 114, 115) (Fig. 3). A global measure of SNA can be assessed from analysis of plasma or urine catecholamine concentrations. Norepinephrine is the primary neurotransmitter released from postganglionic sympathetic nerves in response to neural firing and has been mainly used to estimate sympathetic activation to stressors, particularly exercise; however, epinephrine concentrations have also been utilized. Although useful, a caveat to using plasma concentrations is that they cannot account for the complexities of sympathetic neuronal discharge including reuptake of norepinephrine back into nerve terminals, extraneuronal metabolism of norepinephrine, and/or clearance of norepinephrine. These measures also do not account for potential regional differences. To minimize such concerns, the work of Murray Esler and associates (74, 114, 115) has pioneered the use of radiotracer techniques in humans for the determination of global or specific organ spillover rates of norepinephrine. The general use of these techniques is limited by the required catheterizations, invasiveness, and medical oversight required; however, when utilized important information regarding the control of sympathetic outflow has been garnered, particularly during dynamic exercise (73, 129, 169, 208, 368, 369). Studies have also used spectral analysis of BP variability to estimate peripheral sympathetic drive (248); however, the validity of these indices is questionable (335) and studies relying solely on such measures should be interpreted with caution. In humans, a direct assessment of central sympathetic activity can be made using microneurography to selectively record from postganglionic muscle or skin sympathetic nerves (349). These recordings are analogous to direct recordings of SNA (i.e., lumbar) commonly obtained in animals by the surgical implantation of recording electrodes onto sympathetic fibers however, for obvious reasons, direct sympathetic recordings to internal organs cannot be made in humans. Nevertheless, direct muscle SNA recordings strongly correlate with renal, cardiac and whole-body noradrenaline spillover (368, 369). In addition, the temporal resolution that can be achieved with direct measurements from sympathetic nerves is superior to all other sympathetic measurement techniques. Overall, much of the information regarding the sympathetic nervous system and its regulation in humans has been determined using microneurographic recordings and many of the exercise studies will be highlighted in this review. Peripheral sympathetic activity at exercise onset During large muscle mass dynamic exercise direct microneurographic recordings of SNA in humans are challenging due to contamination from action potentials generated by skeletal muscle cells, as well as movement and the associated risk of recording electrode displacement from the nerve. Thus, much of the information regarding SNA onset dynamics in humans has been obtained from inactive limbs using smaller muscle masses (e.g., handgrip) as well as from animal studies where the electrode can be firmly implanted around the nerve. Allyn Mark and Gunnar Wallin obtained some of the first direct recordings of muscle SNA in exercising humans from the peroneal nerve of the inactive leg during isometric Volume 5, April

25 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology Cardiac NA spillover Cardiac sympathetic nerve scanning: MIBG, 18 F dopamine T L 1 2 NA plasma concentration (arterial and venous): forearm NA spillover Renal NA spillover 2s 3 Clinical microneurography Skeletal muscle Skin Figure 3 Summary of methods used for measuring regional SNA in humans. Sympathetic nerve firing can be measured directly in postganglionic sympathetic nerve fibers of skin and skeletal muscle using the technique of microneurography. Isotope dilution methods can be used to measure the spillover rates of norepinephrine to plasma from individual organs, which provides an assessment of regional SNA in the limbs as well as specific organs. Cardiac sympathetic nerve scans can be used to study the anatomy of the sympathetic innervation of the heart. Reprinted, with permission, from (74). handgrip (198). This study set forth the concept that muscle SNA does not increase at the onset of exercise, but rather there is a delay of about 1 min during an isometric handgrip contraction performed at 30% MVC (Fig. 4). In contrast, skin SNA increases at the immediate onset of isometric handgrip and leg extension exercise (267, 344, 358, 359, 384). The increase in skin SNA at exercise onset appears proportional to the exercise intensity and remains elevated throughout exercise (267, 283, 344, 358, 359). Likewise, animal investigations demonstrated immediate increases in renal and cardiac SNA at the onset of dynamic exercise (54,237,346). Overall, the onset of dynamic exercise does not appear to produce mass, uniform sympathetic discharge to all tissues, rather the majority of studies suggest selective sympathetic activation at exercise onset. To further explore such differential control and better understand the latency of muscle sympathetic activation during dynamic exercise, Seals and Victor undertook a series of studies using rhythmic handgrip and one or two-arm cycling. In general, these studies further supported the idea that during dynamic exercise sympathetic outflow to skeletal muscle did not immediately increase above resting values and there was a latency of approximately 30 to 60 s before a significant change occurred, which was most notable at higher workloads or when the circulation to the exercising muscle was occluded (299, 300, 353, 354). Around the same time, Saito and colleagues also performed several studies examining SNA responses and latencies to different modes of exercise ( , 284, 286) and in general, observed a temporal response pattern similar to that reported by Seals and Victor. Aside from identifying the temporal pattern of muscle sympathetic activation during exercise, these studies also indicated that the buildup of metabolites and stimulation of metabolically sensitive skeletal muscle afferents contribute importantly to the muscle SNA response to exercise. The specific contributions of the different neural mechanisms to exercise mediated sympatho-excitation will be discussed in detail in the following sections. 482 Volume 5, April 2015

26 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans Integrated muscle SNA neurogram Control Handgrip First min Handgrip Second min PEI Recovery MAP (mmhg) HR (bpm) Figure 4 Muscle SNA, mean arterial pressure (MAP), and HR responses to isometric handgrip (35% MVC) followed by a period of post exercise ischemia (PEI) to isolate the muscle metaboreflex. As originally reported by Mark et al. (198) HR increases from the onset of exercise, MAP rises more gradually and a delay is present for muscle SNA consistent with muscle metaboreflex mediation. This is supported by the maintenance of muscle SNA and MAP during PEI, a period in which muscle metaboreflex mediated responses are isolated from central command and muscle mechanoreceptors. In contrast, HR returns to baseline values during PEI. These concepts were further substantiated by studies examining muscle SNA in the arm (e.g., radial nerve) during dynamic leg exercise (30, 266, 286). A reduction in muscle SNA was reported during the preparation for (30) and at the onset of low intensity leg cycling exercise (30, 286); however, this sympathetic inhibition was overcome at higher exercise intensities (>60% peak workload) where increases in muscle SNA were observed. Ray et al. (266) reported that during one legged kicking muscle SNA was reduced below baseline when exercise was performed in the upright position, and unchanged during supine exercise. As discussed in more detail below, these findings point to the important modulatory role played by venous return, central blood volume, and thus cardiopulmonary baroreceptor loading status to the muscle SNA responses at the onset of low-intensity dynamic exercise, but at higher exercise intensities this sympatho-inhibitory effect is obscured by other sympatho-excitatory mechanisms (e.g., exercise pressor reflex and central command). Collectively, studies examining SNA responses at the onset of exercise support a nonuniform sympatho-excitation that likely assists in initiating the appropriate cardiovascular adjustments to exercise. In this regard, the reported immediate increases in SNA to the kidney and skin would be important for redistribution of blood flow to active muscles upon the initiation of exercise. At the same time, lack of an increase or an inhibition of muscle SNA would facilitate vasodilation and increased blood flow within active skeletal muscles. Lastly, the increase in cardiac SNA would elevate HR (along with reduced parasympathetic nerve activity) and cardiac contractility contributing to increases in cardiac output, which would be preferentially distributed to active skeletal muscle due to the aforementioned peripheral SNA responses. While this differential sympathetic activation pattern at the onset of exercise seems reasonable and appropriate, it is important to note that some studies have indicated an immediate and sustained exercise-induced increase in lumbar SNA to skeletal muscle in conscious rats (56) and unanesthetized decorticate cats (121). Peripheral sympathetic activity during steady-state exercise In contrast to the onset of exercise, there is substantial evidence that SNA increases in an intensity dependent manner during sustained levels of dynamic exercise in humans. Indeed, for the most part, studies using direct muscle SNA recordings, plasma norepinephrine as well as norepinephrine spillover have all suggested that sympathetic activation Volume 5, April

27 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology becomes progressively greater as exercise intensity and duration is increased (24,103,146,186). Galbo et al. (103) reported a progressive increase in venous norepinephrine concentration during graded treadmill exercise at workloads above 50% maximal oxygen uptake (VO 2 max). Likewise, arterial and venous norepinephrine concentrations were found to progressively increase during incremental leg cycling (24, 186, 334, 341). An exercise intensity dependent and marked elevation in whole body norepinephrine spillover was also observed, to which a large contribution was attributable to the kidney (188,341). However, Seals et al. (300) reported a strong relationship between plasma norepinephrine and muscle SNA during graded intensity two-arm cycling, but notably, at low workloads neither norepinephrine nor muscle SNA were increased. In addition, more recent work by Ichinose et al. (146) demonstrated that during 6-min stages of graded leg cycling from very mild to exhaustive exercise an initial decrease in directly recorded muscle SNA at low workloads was followed by progressive increases reaching an approximate 265% elevation from baseline during exhaustive exercise (Fig. 5). These data clearly demonstrate that profound increases in muscle SNA can occur with dynamic exercise. Aside from exercise intensity, the duration of exercise has a major impact on the temporal pattern and degree of sympatho-excitation. In this regard, Davy and colleagues (52) reported an increase in plasma venous norepinephrine after 10 min of treadmill walking at approximately 65% peak oxygen uptake (VO 2 peak) that increased further as exercise continued. Furthermore, progressive increases in norepinephrine have been reported during 3 h of treadmill running at 60% of VO 2 peak (103). It is important to note that both the intensity and duration of exercise will combine to dictate the overall sympathetic response. For example, during 30 min of sustained leg cycling performed at 25% of maximum workload, plasma norepinephrine concentration and wholebody norepinephrine spillover were increased by 15 min of exercise and then plateaued to remain elevated throughout (188). In contrast, 30 minutes of exercise performed at 65% of maximum workload evoked a much larger increase in plasma norepinephrine concentration and whole-body norepinephrine spillover that progressively increased throughout the exercise period. A similar progressive increase in sympathetic activation was observed using direct muscle SNA recordings obtained during 30 min of upright cycling at 40% VO 2 max (285). Thus, in describing the sympathetic response to dynamic exercise the duration and intensity need to be considered as both contribute importantly to the degree of sympathetic activation. An important distinction that requires consideration is whether sympathetic responses are being measured in active or inactive skeletal muscle beds. Indeed, many of the studies examining sympathetic responses to exercise in humans, particularly those using direct recording of muscle SNA, have assessed the inactive limb. Furthermore, whole body norepinephrine measures represent a marker combining active and inactive beds. This becomes quite important MSNA MSNA MSNA MSNA MSNA MSNA MSNA 250 BP BP BP BP BP BP BP Rest Very mild Mild Moderate Heavy Exhausting Recovery 30 s Figure 5 Muscle SNA responses to a bout of incremental leg cycling ranging in intensity from very mild to exhausting. Raw recordings of arterial BP and muscle SNA (MSNA) during rest, very mild, mild, moderate, heavy, and exhausting exercise and recovery in a representative subject from Ichinose et al. (146). After an initial decrease in MSNA from rest during very mild exercise, MSNA was increased progressively as exercise intensity increased. Reprinted, with permission, from (146). 484 Volume 5, April 2015

28 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans because the limited data available suggests that the active limbs may be the major targets of the increase in sympathetic activation. For example, it has been reported that 60% of norepinephrine release during submaximal dynamic exercise comes from the skeletal muscle circulation (73). In addition, the classic work of Savard and colleagues (291) demonstrated that norepinephrine spillover was greater from the active leg compared to the inactive during one legged knee extension exercise undertaken at 50% and 100% of maximum workload. Although the concept that there is greater SNA to active skeletal muscle is intriguing and seems reasonable as this may be needed to offset local metabolically mediated vasodilation, not all studies have reported a difference between active and inactive SNA during exercise. In this regard, Hansen et al. (124) reported similar increases in directly measured muscle SNA in the active and inactive limb during ischemic unilateral static toe extension at 20% MVC. The reason for the lack of a difference in muscle SNA between the exercising and nonexercising limbs in this study is not clear but may relate to the exercise modality employed, the low exercise intensity, and/or the small muscle mass engaged. Indeed, although ischemic, a mild intensity was used in this study, and when matched for relative exercise intensity, sympathetic activation will be greater when a larger compared to smaller muscle mass is engaged (23). However, even though the size of the contracting muscle mass reportedly increases the magnitude of the resultant SNA and BP response, this association is complex and likely affected by a multitude of factors including but not limited to skeletal muscle fiber type, exercise mode, and local blood flow (229, 297), as reviewed elsewhere (88). In summary, the existing literature provides clear evidence for intensity and duration-dependent increases in muscle SNA to both the active and inactive limbs during isometric and dynamic exercise in humans. This is accompanied by increased norepinephrine spillover from the cardiac (129,368), renal (129,341), and splanchnic (208) vasculature. Nevertheless, regional differences in the temporal pattern of the sympathetic activation and the impact that both intensity and duration have on the sympathetic response have to be carefully considered when examining studies reporting SNA during exercise. In the following sections of this review, we will discuss the neural mechanisms underpinning the parasympathetic and sympathetic adjustments to exercise. Neural Cardiovascular Control Mechanisms Central command In 1886, Zuntz and Geppert proposed the existence of a cortical control mechanism that simultaneously activated respiratory centers and voluntary locomotor pathways. This feed forward parallel activation of locomotor and respiratory neural circuits was further advanced to incorporate the cardiovascular responses to exercise (156, 180, 181). Originally termed cortical irradiation (181) and later central command (113), this concept refers to descending neural signals that elicit skeletal muscle contraction and concomitantly activate central nervous system centers involved in the control of autonomic neural outflow to the cardiorespiratory system. It is now clear that an individual s perception of effort contributes to the magnitude of central command during exercise, independent of the actual force produced (Fig. 6). However, it is important to note that effort sense is a complex variable that can be influenced by numerous stimuli (e.g., pain), which provide afferent feedback to cardioregulatory centers of the brain (377, 378). Thus, in addition to the traditional feed-forward concept (180, 181, 390), it is likely that central command also involves feedback mechanisms (377, 378), and in this regard multiple brain sites (116). Due to the complexity of central command, the identification of specific brain region(s) responsible for evoking autonomic adjustments has remained elusive. However, putative locations for central command have been identified in animals using direct electrical or chemical stimulation of neural structures including the Fields of Forel, motor cortex, insular cortex, mesencephalic, and hypothalamic areas (1, 15, 68, 314, 332, 362). The advancement of neuroimaging techniques has provided an opportunity for translation of these findings to humans and the insular and anterior cingulate cortices have been suggested as human brain regions that are activated by central command during exercise (45, 168, 234, 235, 339, 379, 380). More recently, direct electrical stimulation of midbrain areas during neurosurgery (e.g., deep brain stimulation) in awake humans has been used to enhance our understanding of potential central command areas (16, 117, 118, 338) (Fig. 7). While constrained by electrode placement and a patient population (e.g., chronic pain), these novel studies have indicated that the thalamus, subthalamic nucleus, substantia nigra, periaqueductal gray, and periventricular gray may all be involved in determining the cardiovascular response to exercise. Collectively, although these studies identify specific regions of interest, the development of an integrated neurocircuitry model of central command remains an open area of investigation. The influence of the exercise phase, duration, intensity, and modality on this complex neural mechanism is discussed below. Central command at exercise onset The idea of a central neural signal driving cardiovascular responses during exercise arose from early observations that HR increased in anticipation of and immediately at the onset of exercise a response that was suggested to be too rapid to be explained by reflex mechanisms in the contracting skeletal muscle (29, 108, 156, 180, 181). These initial observations, intuitively, albeit indirectly, suggested a fast feed-forward neural mechanism, and were supported by subsequent studies demonstrating an increase in HR within the first beat during static arm contraction (143) and large muscle dynamic exercise (153, 231, 382). A more definitive role for central command in the initial HR response to exercise has been Volume 5, April

29 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology 100 Actual exercise 100 Imagined exercise 95 High hypnotizability Low hypnotizability 95 Heart rate (beats/min) Mean blood pressure (mmhg) Perceived exertion (units) Rest Time (min) Rest Time (min) Figure 6 Imagined and actual handgrip exercise elicits similar cardiovascular responses. Heart rate, mean arterial blood pressure, and rating of perceived exertion during actual (left) and imagined (right) handgrip exercise. Subjects were categorized into low (n = 4) and high (n = 5) hypnotizability groups. Muscle electromyographic recordings demonstrated no measurable increases in force during imagined handgrip, not shown. The data are presented as mean ± SD. P < 0.05 low versus high hypnotizability. Reprinted, with permission, from (379). 486 Volume 5, April 2015

30 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans (A) (B) LFP (μv) (C) 1000 PSD (μv 2 /Hz) Time (s) Rest Anticipation Exercise Recovery (D) Normalised power ratio Hz 8-12 Hz Hz Frequency band Frequency (Hz) Hz Hz Figure 7 Electrophysiology recordings from the periaqueductal gray suggest that this neural region is involved in mediating the cardiovascular adjustments to exercise. Three patients had stimulation electrodes placed in the periaqueductal gray for the treatment of chronic neuropathic pain. Local field potentials (LFP) were collected from the periaqueductal gray during resting conditions, anticipation of exercise, cycling at 15 W (30-60 s) and recovery from exercise. (A) Original LFP recordings from the periaqueductal gray. (B) Magnetic resonance image illustrating electrode placement in the periaqueductal gray. (C) Mean power spectral density from all three subjects. (D) Normalized spectral changes (rest = 1.0) divided into frequency bands. P < 0.05, P < 0.01, P < versus rest. Reprinted, with permission, from (118). provided from studies using passive cycling exercise (382) and partial motor paralysis (153, 301) to alter the level of central command influence. According to the latter approach, partial neuromuscular blockade decreases the contractile ability of the paralyzed muscles and thus, a greater neurally generated signal is required to maintain the same level of absolute force (i.e., exaggerated central command input). In this regard, the initial rise in HR in response to voluntary contractions is unaffected by partial neuromuscular blockade (i.e., same HR response with less absolute muscular force developed), suggesting that the rapid tachycardic response is related to voluntary effort, rather than feedback from the contracting skeletal muscle (153, 301). These findings are in line with classic work in cats in which stimulation of locomotor brain regions elicited stimulation-dependent increases in cardiovascular variables, despite lack of muscle contraction due to deep anesthesia or paralysis (57, 67, 363). Thus, it is well accepted that the initial rapid increase in HR at the onset of exercise is due to descending central nervous system input, although as discussed later, reflex inputs contribute quickly. Due to the latency of parasympathetic and sympathetic influences on HR, the general assumption has been that central command-induced cardiac vagal withdrawal mediates the increase in HR at the onset of exercise. Indeed, vagal blockade with atropine blunts the rapid rise in heart in response to static contractions (98,108,143). However, recent evidence suggests a potential role for early increases in cardiac sympathetic outflow during spontaneous motor activity in a feline model (346), although the translational applicability of these findings to humans remains yet to be determined. In contrast to cardiac autonomic control, relatively few investigations have examined a direct role for central command mediated peripheral vascular regulation at the onset of exercise. However, BP increases in anticipation of exercise, relative to the perceived level of effort necessary, which is consistent with a centrally mediated mechanism (116). Using brief (3 s) maximal isometric contractions, Iwamoto et al. (153) demonstrated that the exercise induced increase in BP was attenuated, but not eliminated, following partial neuromuscular blockade. That is, central command, in the absence of reflex skeletal muscle feedback, contributed in part to the rise in BP. These findings are consistent with work in cats in Volume 5, April

31 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology which pharmacological blockade of skeletal muscle stretch activated ion channels did not influence early increases in BP during spontaneous muscle contractions (i.e., central command induced pressor response) (201). However, the autonomic contribution to potential central command induced early increases in BP has yet to be fully determined, although in support of a sympathetic contribution, BP and renal SNA have been shown to increase prior to and at the onset of static exercise (206), locomotion (277), grooming (203) and treadmill exercise (237) in animals. Although not specifically designed to examine the onset of exercise, work from Victor and colleagues using brief rhythmic handgrip exercise and partial neuromuscular blockade lends some additional insight (355). Short 3-s handgrip contractions below 50% MVC were shown to have no influence on muscle SNA. However, under intense conditions (75% MVC) every muscular contraction was accompanied by a synchronized burst of muscle SNA and this synchronization persisted following partial curization, although muscular force development was minimized. These observations highlight the potential for central command to increase muscle SNA at the onset of exercise, albeit likely only at high intensities. In contrast, central command appears crucial for controlling skin SNA, due to the rapid increases observed prior to and at the immediate onset of exercise (267, 344, 358, 359, 384). At the onset of dynamic muscular contractions a rapid increase in active skeletal muscle vasodilatation occurs, concomitant with a decrease, increase or no change in BP depending on the exercise modality and intensity (174). While the cause of this rapid vasodilation remains unresolved and has been attributed to a number of factors including the muscle pump and vascular compression, vasodilatory or metabolic factors, myogenic responsiveness, and sympathetic vasoconstrictor withdrawal, the concept of a sympathetically mediated vasodilatory mechanism (e.g., cholinergic) has remained contentious in humans. Cholinergic (muscarinic) blockade has been shown not to influence the hyperemic response at the onset of rhythmic handgrip exercise (308); findings that are consistent with the observation that there are no obvious sympathetic cholinergic vasodilator fibers in human skeletal muscle (25,348). Furthermore, the increase in forearm blood flow in response to an attempted maximal contraction is minimal following limb paralysis with pipecuronium, an agent that blocks postsynaptic nicotinic receptors with no effect on acetylcholine release, suggesting that acetylcholine spillover from other areas (e.g., motor nerves) is not involved in skeletal muscle vasodilation in humans (63, 372). In contrast, Secher and colleagues (303) demonstrated that BP was stable at the onset (first 6 s) of a control bout of low-intensity dynamic exercise, but with partial curarization BP decreased, suggesting that the augmentation of central command stimulates a peripheral neurogenic vasodilatory mechanism. More recent examinations using a novel technique of motor imagery of exercise to stimulate central command have suggested that a neural signal induces vasodilation in skeletal muscle (152). However, the transient increase in lower limb vascular conductance evoked by an isometric contraction of the contralateral calf muscle is no different when performed either volitionally or electrically evoked, suggesting that central command is not requisite for this rapid vasodilatory response (89). Despite conflicting findings in humans, a number of observations in animal models demonstrate that stimulation of certain neural regions can produce skeletal muscle vasodilation (1, 15, 70, 142, 202, 204) and that the increases in blood flow and vascular conductance at the onset of muscular contractions are blocked to a similar extent with either ganglionic blockade or atropine (177). Overall, while a role for central command in mediating the peripheral vascular responses at the onset of exercise has been demonstrated in humans, further investigations are warranted, including considerations for the specific autonomic contributions relative to the exercise modality, muscle mass, and intensity. Central command during steady-state exercise As highlighted above, a primary strategy used to study the potential role of central command during exercise in humans has been neuromuscular blockade. Freychuss (97) reported that an unsuccessful attempt to isometrically contract the forearm muscles, following complete neuromuscular blockade, was still accompanied by an increase in HR and BP, albeit approximately 50% of the normal response. In line with this, partial neuromuscular blockade to augment the level of central command has generally resulted in an exaggerated cardiovascular response during static and dynamic exercise (13, 106, 187, 224, 250, 261, 351) (Fig. 8). Although such a relationship is not as clear during high intensity isometric contractions (107, 184) and maximal dynamic exercise (104). Studies employing additional strategies to modulate central command input such as electrical stimulation of resting skeletal muscle (180), hypnosis (49, 339, 370, 379, 380), muscle tendon vibration (113, 155, 246), and patients with unilateral limb weakness (151, 385) or sensory neuropathies (61) have also demonstrated a role for central command in mediating the cardiac acceleration and increases in BP in response to exercise. For example, Williamson et al. (379) showed that the HR and BP responses to actual and imagined (i.e., hypnosis with no force produced) handgrip exercise were identical, supporting the concept that central command mediated responses are dictated by the perception of effort (Fig. 6). Although less extensively investigated, the increase in cardiac output during exercise appears to be mediated in part by central command, primarily due to modulation of HR (192,250,326,383), although discrepant findings have been reported (13, 151). The traditional concept has been that the central command-induced increase in HR during exercise is due to parasympathetic withdrawal. Indeed, muscarinic, but not nonselective β-adrenergic, blockade, reduces the central command-induced tachycardia during exercise with partial curization (224, 354). However, this area may warrant further investigation in humans given recent findings that central command has been shown to primarily influence cardiac 488 Volume 5, April 2015

32 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans Blood pressure (mmhg) Heart rate (beats/min) Time (min) Figure 8 Augmenting the level of central command with partial neuromuscular blockade exaggerates the cardiovascular responses to static handgrip exercise. Mean arterial blood pressure and heart rate at rest, during a 2 min handgrip contraction at 15% maximal voluntary contraction (dashed lines) and for 2 min of recovery. The data are presented as the mean from 12 healthy subjects (5 female). control, contraction with tubocurarine that the subjects were able to maintain for 2 min, contraction with tubocurarine that the subjects were not able to maintain throughout the 2 min exercise period. P < 0.05 resting and exercise values during tubocurarine are different from control. Reprinted, with permission, from (224). sympathetic nerve activation (345, 346) versus parasympathetic nerve activity (160) in cats. In regards to BP, Mitchell et al. (224) demonstrated that α- or β-adrenergic blockers alone or in combination with muscarinic blockade do not block the central command induced elevation in exercising BP during partial neuromuscular blockade. These findings illustrate a redundancy in the autonomic control of BP via central command signals, although additional investigations are needed. Exaggerated elevations in plasma catecholamine concentrations have been found when central command is increased during static (250) and dynamic (104) exercise with partial neuromuscular blockade, implying a role for central command mediated sympathetic activation. Indeed, elevations in skin SNA during handgrip exercise with neuromuscular blockade were shown to be similar to the responses during a control contraction despite the decrease in force development and feedback from skeletal muscle afferents (358). These findings highlight the importance of central command in mediating skin SNA responses to exercise. In contrast, as mentioned above, using direct recordings of muscle SNA in experiments designed to isolate the reflex effects of skeletal muscle metaboreceptor afferents, Mark and colleagues suggested that central command has a minimal influence on sympathetic outflow to the skeletal muscle vasculature during low- to moderate-intensity static and dynamic exercise (198, 354). Furthermore, the magnitude of the rise in muscle SNA during contractions following partial neuromuscular blockade is minimal (351). Collectively, these findings suggest that central command is crucial for increasing efferent skin SNA during exercise, but has a modest effect on skeletal muscle sympathetic outflow at low to moderate exercise levels, although a more prominent role may be apparent during high intensities of exercise (355). Exercise pressor reflex The contribution of a reflex originating in skeletal muscle to the cardiorespiratory response to exercise was recognized toward the end of the 19 th century (156, 390); however, it was the work of Alam and Smirk (5, 59) that unequivocally established the importance of afferent feedback from skeletal muscle in evoking the pressor response to exercise in humans (Fig. 9). Since this seminal work, an abundance of research has been devoted to understanding this neural reflex mechanism emanating from skeletal muscle, including the identification of its afferent and efferent components, which have collectively been termed the exercise pressor reflex (209). During exercise both mechanically and metabolically sensitive sensory fibers provide feedback via the dorsal horn of the spinal cord to brainstem cardiovascular areas in response to mechanical and metabolic stimuli, respectively (41, 163, 165, 209). The muscle mechanoreflex is mainly comprised of thinlymyelinated group III afferent neurons whose receptors are primarily activated by mechanical deformation as induced by changes in pressure or stretch (318, 381). Thus, the muscle mechanoreflex is primarily activated at the immediate onset of muscle contraction coincident with a distortion of the mechanoreceptors receptive field. Indeed, group III afferents have been shown to fire with an abrupt and profound burst of impulses at the onset of muscle contraction (162, 219). In contrast, the afferent fibers of the muscle metaboreflex are mostly unmyelinated group IV neurons whose receptors are primarily chemically sensitive and stimulated by metabolites produced by contracting skeletal muscle (163, 222). Therefore, the muscle metaboreflex is thought to require a sufficient period of time before it responds due to the delay in production of metabolites by contracting skeletal muscle (198). However, this separation in activation of these exercise pressor reflex afferent neurons is not absolute as group III and IV fibers exhibit polymodal qualities such that some group III fibers respond to metabolic changes while some group IV fibers respond to mechanical distortion (219). In addition, both fiber types can be sensitized (191, 275) or desensitized (275) by changes in the metabolic milieu of the muscle Volume 5, April

33 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology (A) interstitium. Overall, there is substantial research demonstrating that during muscle contraction, activation of both mechanically and metabolically sensitive afferent fibers contribute importantly to the autonomic responses to exercise and as such, play a major role in mediating the neural cardiovascular adjustments to exercise. (B) Systolic pressure (mmhg) Systolic pressure (mmhg) Exercise Exercise Cuff ischaemia From Alam and Smirk 1937 Recovery Recovery Time (min) Figure 9 Original report identifying a BP-raising reflex originating in skeletal muscle in humans. (A) Schematic showing preparation for the experimental protocol used in which a cuff was placed around the exercising forearm to perform ischemic exercise, while a cuff on the opposite arm was used to measure arterial BP. (B) Shows part of the seminal results from this study documenting the mild increase in BP during freely perfused rhythmic exercise (top panel) in comparison to the massive increase in systolic BP during and after exercise when the forearm was made ischemic by cuff inflation to supra-systolic pressure. Reprinted, with permission, from (220). Skeletal muscle mechanoreflex at exercise onset As noted in the previous section, there is substantial work in humans attributing the immediate HR increase at the onset of exercise (e.g., the shortening of the first R-R interval) to a central command mediated reduction in cardiac parasympathetic activity (153, 181, 301, 382). However, pronounced increases in HR at the onset of muscular contraction can be induced in the absence of central command. Krogh and Lindhard (180) demonstrated in 1917 that percutaneous electrical stimulation can evoke a muscular contraction accompanied by an increase in HR, but only after the first R-R interval. As central command is bypassed during an electrically stimulated contraction the resultant HR response is attributable to muscle afferent feedback. Although, a muscle-heart reflex was proposed by Hollander and Bouman (143) with a latency of 550 ms during either voluntary or electrically stimulated exercise, these investigators may have inadvertently stimulated the skeletal muscle afferents directly (153). Isometric contractions evoked by ventral root stimulation in anesthetized rats has also been shown to produce a rapid decrease in cardiac parasympathetic activity (213) and increase in sympathetic activity (205, 345, 346). al-ani et al. (6) reported that when cardiac parasympathetic activity is increased during expiration the magnitude of the HR response to electrically evoked upper arm flexion is greater than when contractions are performed during inspiration, thus indicating that feedback from skeletal muscle afferents can inhibit cardiac parasympathetic tone. Mechanically sensitive group III skeletal muscle afferents are known to be robustly activated at the onset of an electrically evoked muscular contraction (163), and as such the rapidity of the cardiac autonomic responses described above suggests a predominant role for these afferents. However, the use of electrically evoked contractions does not permit the relative contribution of mechanically and metabolically sensitive muscle afferents to be dissected. Passive limb movement and external compression have been employed to experimentally activate mechanically sensitive muscle afferents, without engaging those that are metabolically activated. The relative strengths and weaknesses of the approaches used to investigate the influence of human muscle mechanoreceptors on cardiac autonomic control have been reviewed in detail elsewhere (163). Passive hindlimb stretch in cats evokes an increase in HR associated with transient cardiac sympathetic nerve activation and a more sustained reduction in parasympathetic activity (205, 226). Studies in humans utilizing medical anti-shock trousers to compress the limbs (381) and passive muscle stretch (17) have reported no effect on HR. However, HR was shown to increase 490 Volume 5, April 2015

34 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans R-R interval change (ms) Stretch alone Stretch with glycopyrrolate Time (s) Figure 10 Change in R-R interval evoked by passive calf stretch under control conditions (black bars) and with glycopyrrolate (gray bars). The shortening of the mean (+SE) R-R interval in response to stretch was significantly ( P < 0.05) attenuated with cholinergic (muscarinic) blockade with glycopyrrolate in 3 subjects. Reprinted, with permission, from (110). during 4 s of passive leg cycling movements (231), while Williamson et al. (382) demonstrated that when passive leg cycling movements were combined with percutaneous electrical stimulation and triggered early enough in the cardiac cycle (first one-third) an instantaneous increase in HR was elicited. Subsequent to this Coote and colleagues (109, 110) demonstrated that passive calf muscle stretch evoked a transient HR increase that was accompanied by a reduction in HR variability and abolished with prior administration of glycopyrrolate (Fig. 10). Thus, despite some evidence to the contrary, it appears that activation of mechanically sensitive muscle afferents in humans can increase HR via inhibition of cardiac parasympathetic activity. However, it is pertinent to note that of the mechanically sensitive group III muscle afferents that respond to tendon stretch, only 50% have been shown to respond during an isometric muscle contraction (132). The influence of the muscle mechanoreflex on peripheral SNA is arguably less well understood than its influence on cardiac autonomic control. Animal studies have identified that the muscle mechanoreflex increases renal (175, 352) and muscle (140) SNA. In humans, Herr et al. (135) reported that muscle SNA increases at the onset of isometric quadriceps contractions at 25% maximum voluntary contraction with a latency of 4-6 seconds. While the short latency of this response is indicative of a muscle mechanoreflex effect, a contribution from central command and/or the muscle metaboreflex cannot be excluded. Isolated muscle mechanoreflex activation in humans, induced by passive dynamic forearm stretch, has been reported to evoke a slight transient increase in muscle SNA with a latency of 1 to 3 s (46). Thus, it appears that the muscle mechanoreflex evokes rapid, albeit modest, autonomic alterations at the level of both the heart and skeletal muscle vasculature, and thus contributes to the initial cardiac and hemodynamic responses to exercise in healthy humans. Skeletal muscle mechanoreflex during steady-state exercise Following an initial burst of activity at the onset of a tetanic contraction the discharge rate of mechanically sensitive group III skeletal muscle afferents typically decreases if the contraction is nonfatiguing (163, 217). However, the discharge rate of these afferents rises or falls with a respective increase or decrease in isometric tension, and their discharge can become synchronized to electrically evoked intermittent isometric tetanic contractions (163, 217). Furthermore, during low-intensity dynamic exercise induced by mesencephalic locomotor region stimulation in cats muscle mechanoreceptor activity becomes synchronized with step cycle (2). Such observations highlight the potential importance of muscle mechanoreceptors to autonomic control not only at the onset of exercise, but also during steady-state dynamic exercise at a low intensity. However, studies in humans using lower body positive pressure have suggested that intense mechanoreceptor stimulation is needed to evoke a mechanoreflex mediated increase in muscle SNA (101, 102). Overall, additional research is needed to delineate the contribution of mechanoreflex to muscle SNA, particularly during dynamic exercise. During a fatiguing isometric contraction at a constant tension the activity of muscle mechanoreceptors typically increases, possibly due to the accumulation of metabolic byproducts within the muscle (165). This sensitization effect is also evident from studies showing that mechanoreceptor discharge is potentiated by ischemia or substances such as bradykinin, arachidonic acid, ATP, or cyclooxygenase products (3, 165, 228). Work examining the potential importance of muscle mechanoreceptor sensitivity on autonomic control in humans is limited. Fisher and colleagues (84) investigated whether the HR and BP responses to passive calf stretch were potentiated when performed during postexercise ischemia, a period in which the concentration of metabolites within the muscle is elevated. The magnitude of both the HR and BP response to passive stretch was the same irrespective of whether muscle metabolites were elevated or not. In contrast, Cui et al. (47) reported that when static passive wrist extension stretch was applied during postexercise ischemia following fatiguing handgrip exercise significant increases in muscle SNA and BP were evoked, whereas no change in muscle SNA or BP were observed when wrist extension was conducted under free-flow conditions. Although, no HR response to static passive wrist extension was observed under either condition, the sensitization effect on muscle SNA and BP was subsequently shown to be diminished following cyclooxygenase inhibition (48). The reason for the conflicting findings of Fisher et al. (84) and Cui et al. (47) may relate to differences in muscle group (calf, forearm), exercise mode (nonfatiguing, fatiguing) or method used to induce passive stretch. Collectively, these observations imply that the accumulation of metabolites within the skeletal muscle during steadystate exercise can sensitize the mechanically sensitize skeletal muscle afferents, most likely leading to an increase in muscle Volume 5, April

35 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology (A) Imp/2 s Time (s) (B) Imp/2 s Time (s) Figure 11 Activation of group III and IV skeletal muscle afferents during dynamic exercise induced by stimulation of the mesencephalic locomotor region in cats. Cumulative histograms for 24 group III afferents (A) and 10 group IV afferents (B) before, during, and after dynamic exercise. The exercise period is denoted by horizontal bars. These findings demonstrate that low-intensity dynamic exercise stimulated both Group III and IV skeletal muscle afferents. Imp, impulses. Reprinted, with permission, from (2). SNA and BP, while the effect on cardiac autonomic control appears minimal. Skeletal muscle metaboreflex at exercise onset While the thin fiber skeletal muscle afferents respond to mechanical distortion and therefore, typically respond immediately at the onset of muscular contraction, those responsive to metabolic perturbation respond with a longer latency (163). These observations, along with the time delay for the accumulation of most metabolites during exercise, means the contribution of muscle metaboreceptors to autonomic regulation is generally believed to be minimal at the onset of exercise. However, the studies of Kaufman and colleagues have illustrated that both group III and group IV skeletal muscle afferents are engaged during short duration (60 seconds) low-intensity dynamic exercise and that their discharge is coupled to the rhythmical contraction of the working muscles (2, 3) (Fig. 11). These direct afferent recordings suggest that the Group IV afferents have the potential to contribute to the autonomic responses in early exercise, prior to the generation of a substantial metabolic error signal. Nevertheless, central command is traditionally viewed as setting the initial autonomic response to exercise, and if the exercise pressor reflex is involved at exercise onset it would likely be through mechanically sensitive group III afferents. A caveat to this viewpoint is that studies in humans isolating the contribution of muscle metaboreceptors to the initial autonomic responses to exercise are lacking. This likely has to do with the complexity of onset kinetics in regards to neural interactions and the robust nature of the cardiovascular responses with the initiation of exercise. Skeletal muscle metaboreflex during steady-state exercise The first attempt to experimentally isolate the contribution of the skeletal muscle metaboreflex to the cardiovascular response to exercise was made by Alam and Smirk (5). In a landmark study, subjects performed rhythmic forearm and calf contractions, first under free-flow conditions and second with circulation to the limb occluded such that they were ischemic both during and following exercise (see Fig. 9). The postexercise ischemia maneuver trapped the exercise-induced metabolites within the previously active muscle, thus preserving their stimulatory effect on metabolically sensitive skeletal muscle afferents in the absence of the muscle mechanoreflex or central command. This maneuver, first performed with direct muscle SNA measures by Allyn Mark, Gunnar Wallin, and colleagues, has been used in numerous studies to investigate the muscle metaboreflex in humans. Consistently, during postexercise ischemia, the exercise-induced increase in BP and muscle SNA remain robustly elevated (see Fig. 4) (198), whereas the HR response appears to depend 492 Volume 5, April 2015

36 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans significantly on the muscle mass and/or exercise modality studied (4, 96). Indeed, HR returns toward baseline levels during postexercise ischemia following rhythmic or isometric handgrip (87, 198, 230), static leg extension (149), and isometric calf plantar flexion (89). In contrast, HR is elevated above baseline during postexercise ischemia following rhythmic calf plantar flexion (4) and cycling (96, 128) with both legs. Such findings imply that the autonomic control of the heart is differentially modified during these maneuvers. In addition, the intensity of the preceding exercise and thus, magnitude of metaboreceptor activation during postexercise ischemia can also play a role in the degree to which HR is affected by the muscle metaboreflex. This is likely due to the intensity dependence of cardiac sympathetic activation via the metabolically sensitive afferents (87). As the muscle metaboreflex activation is a potent stimulus of the sympathetic nervous system it is seemingly incongruous that in many circumstances HR remains at baseline values during postexercise ischemia (87,149,198,230). Part of the explanation for this seems to be the loss of the powerful inhibitory input from central command (224) and muscle mechanoreflex (110) to cardiac parasympathetic preganglionic nuclei upon the cessation of exercise. In addition, the excitatory input to these nuclei is likely increased as a consequence of arterial baroreceptor stimulation resulting from the robust elevation in BP during postexercise ischemia. As originally demonstrated in dogs (238), and more recently in humans (87), the activation of cardiac parasympathetic activity at this time can overpower a potential sympathetically mediated elevation in HR. Indeed, O Leary (238) reported that during postexercise ischemia following treadmill exercise in canines HR returned toward baseline levels under control (no drug) conditions, whereas with the administration of atropine to block cardiac parasympathetic activity HR remained at the level observed during exercise. Thus with the cardiac parasympathetic activity removed a sympathetically mediated elevation in HR elicited by the muscle metaboreflex was seemingly revealed. In humans, cholinergic muscarinic blockade with glycopyrrolate unmasked an elevation in HR during postexercise ischemia following low intensity handgrip (87). This direct pharmacological evidence (Fig. 12) supported earlier work showing increases in HR variability derived indices of both cardiac parasympathetic (230) and sympathetic (149) nerve activity during postexercise ischemia in humans. As mentioned above, an elevation in HR with postexercise ischemia is observed following some modes of exercise. We have reported a modest HR elevation during postexercise ischemia subsequent to moderate intensity handgrip (40% maximum voluntary contraction for 2 min) that was eliminated with β-adrenergic blockade (87) (Fig. 12). (A) Heart rate (beats/min) (B) Δ Heart rate (beats/min) % MVC Rest IHG PEI-M Recovery Rest IHG PEI-H Recovery IHG 360 Time (s) Control Beta-adrenergic blockade Parasympathetic blockade PEI-M IHG 40% MVC 360 Time (s) Control Beta-adrenergic blockade Parasympathetic blockade Drug P < Phase P < Trail P < Interaction P < PEI-M Figure 12 HR responses to isometric handgrip (IHG) and postexercise ischemia (PEI) under control conditions (black symbols), and following β-adrenergic blockade (light gray symbols) and parasympathetic blockade (dark gray symbols). HR during all experimental phases (A) and change ( ) in HR from rest (B) are shown. PEI-M, PEI following 25% IHG; PEI-H, PEI following 40% IHG. P < 0.05 versus exercise, P < 0.05 versus control, P < 0.05 versus β blockade, #P < 0.05 versus 25% MVC. Reprinted, with permission, from (87) Volume 5, April

37 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology Thus, it appears that strong activation of the muscle metaboreflex may increase cardiac SNA to an extent that can overcome a reactivation of cardiac parasympathetic activity. Intriguingly, HR variability analysis indicates that elevations in HR during postexercise ischemia following leg cycling are associated with reductions in cardiac parasympathetic activity (128). However, pharmacological investigations have shown that this HR elevation persists with either muscarinic or β-adrenergic blockade, implying a redundancy in the sympathetic and parasympathetic control of the heart under these conditions (83). The role of the muscle metaboreflex in cardiac autonomic control has also been evaluated by the occlusion or partial occlusion of perfusion to the exercising skeletal muscles. Unlike postexercise ischemia, however, this maneuver engages the muscle metaboreflex whilst central command and muscle mechanoreceptors are also activated. Experimental hypoperfusion in exercising dogs (238) and humans (83) can evoke a robust increase in HR that is attenuated by β- adrenergic blockade and not affected by cardiac parasympathetic blockade. However, as β-adrenergic blockade does not entirely block this response (83, 238) and as HR variability derived estimates of cardiac parasympathetic activity are reduced during leg cycling with restricted flow (128), the ability of the muscle metaboreflex to inhibit cardiac parasympathetic activity cannot be entirely ruled out. Intriguingly, increases in HR during electrically evoked cycling exercise in individuals with spinal cord injury (no afferent feedback) are attenuated by the inflation of thigh cuffs to restrict blood flow to and from the active limbs (173). This perhaps indicates a role for blood borne factors released from active skeletal muscle on the HR responses to exercise. Rather than the augmentation of skeletal muscle afferent activation during exercise an alternative approach is to pharmacologically block or inhibit their activity. Administration of an epidural anesthetic to partially block skeletal muscle afferent feedback has been used in a number of investigations (82, 95, 99, 223, 302, 316). While a diminished HR and BP response to exercise can result (223), indicative of the importance of skeletal muscle afferents to the normal cardiovascular response to exercise, this has not been a consistent finding, possibly due to variations in the depth of analgesia and confounding effects on efferent neuromuscular control (302). Likewise, this may have to do with the mode of exercise. Although epidural anesthesia caused clear reductions in HR and BP responses to static handgrip (223), a blunting of the BP but not the HR response has been shown during dynamic exercise (172, 326). An important caveat to epidural anesthesia usage is the muscle weakness induced likely means that central motor drive and thus central command are enhanced to produce a fixed workload. However, this limitation may be circumvented by the pharmacological agonism of spinal opioid receptors, thus modulating the ascending activity of skeletal muscle afferent feedback without affecting central motor drive (386). In exercising dogs, intrathecal administration of the opiate agonist morphine virtually abolishes the HR and BP responses to unilateral iliac arterial occlusion (253), while agonism of spinal opioid receptors also attenuates the exercise reflex pressor in cats (141). In humans it has been reported that HR and BP is reduced during leg cycling at moderate-to-high workloads following lumbar intrathecal administration of fentanyl, a morphine analogue and selective μ-opioid receptor agonist, to partially block lower limb muscle afferents (7, 247). Similar results were recently reported during single leg knee extensor exercise following fentanyl administration (8). However, the autonomic basis for such changes remains to be determined. Nevertheless, the available evidence indicates that the major effect of metabolically skeletal muscle afferents on HR occurs via an increase in cardiac SNA whereas the effect on cardiac parasympathetic activity is more minor (83, 87, 149, 230, 238). Unlike the more subtle influence the metaboreflex has on cardiac SNA, there is unequivocal data to support the robust effect metaboreceptors have in increasing muscle SNA. Numerous studies using postexercise ischemia to isolate the muscle metaboreflex or experimental restriction of active muscle perfusion to augment muscle metaboreflex activation, have demonstrated large intensity-dependent increases in muscle SNA and BP. However, the specific chemical product(s) that activate metabolically sensitive muscle afferents remains controversial (162, 164, 219, 220, 315). This literature has been reviewed in detail by others, but some discussion is warranted due to the robust nature in which the metaboreflex stimulates muscle SNA. Several of the metabolites produced by muscular work that have been targeted as candidates for the activation of the muscle metaboreflex during exercise in humans include, but are not limited to lactic acid, potassium, adenosine, arachidonic acid, diprotonated phosphate, prostaglandins, and hydrogen ion (81, 91, 122, 274, 311, 331, 350, 375). However, equivocal results have been reported for most of these candidates with studies supporting and refuting the involvement of each substance to some degree. For example, studies using individuals who produce minimal amounts of lactic acid due to a myophosphorylase enzyme deficiency (i.e., McArdle s disease) have indicated that the BP and muscle SNA responses to static handgrip are markedly attenuated compared to normal healthy subjects (79, 260) suggesting that the production of lactic acid is requisite for the full expression of the muscle metaboreflex (Fig. 13). However, even within this unique patient population disparate results have been reported (356, 357), although it should be noted that Kaufman and colleagues have demonstrated a major role for lactic acid in stimulating skeletal muscle afferents through acidsensing ion channel receptors (ASIC) (131, 210). In regards to the latter, there is an accumulating body of research examining the receptors responsible for activating metabolically sensitive skeletal muscle afferents with a role for transient receptor potential vanilloid 1 receptors, purinergic receptors and the CB1 cannabinoid receptor along with ASIC receptors (120, 190, 376). There is also now research focusing on the central integration of this afferent information. However, 494 Volume 5, April 2015

38 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans (A) Control subject (B) Δ Total MSNA (%) Handgrip time (min) McArdle s patient Handgrip time (min) Δ MAP (mmhg) HR (beats/min) Controls Patients Handgrip duration (% total time) Ischemia recovery Figure 13 Sympathoexcitatory responses to fatiguing static handgrip at 30% MVC in patients with McArdle s disease, who cannot produce lactic acid due to a myophosphorylase deficiency, and age, sex, and bodyweight matched controls. (A) Original muscle SNA (MSNA) record from a patient with McArdle s disease and a control subject with a similar time to fatigue. (B) Summary data showing MSNA, mean arterial pressure (MAP) and HR responses during fatiguing handgrip followed by postexercise forearm ischemia and recovery. The MSNA response to exercise was severely blunted in the McArdle s patients compared to controls. P < 0.05 versus controls. Reprinted, with permission, from (79). these studies are beyond the scope of this review and the interested reader is directed to some excellent recent reviews (162, 164, 219, 220, 315). In an intriguing recent study, Pollak et al. (252) infused metabolites known to be produced during exercise into the abductor pollicis brevis muscle of the hand of healthy humans. Although the sensations of skeletal muscle fatigue and pain were the main outcome variables of this study, rather than cardiovascular or autonomic responses, this approach holds great promise to further probe the metabolite(s) responsible for stimulating skeletal muscle afferents and evoking the exercise pressor reflex mediated cardiovascular response. Interestingly, the infusion of individual metabolites at maximal amounts evoked no fatigue or pain and it was only when a combination of metabolites was infused did the subjects report sensations of fatigue or pain. These findings suggest that it is a combination of substances that excite metabolically sensitive skeletal muscle afferents. Indeed, as the research in this area continues, this is the most probable scenario to evolve in regards to the cardiovascular responses evoked by the muscle metaboreflex, and there is likely also, redundancy that exists among the metabolites capable of stimulating skeletal muscle afferent fibers. Emerging evidence also indicates that metabolically sensitive afferent fibers emanating from respiratory muscles may play a role in increasing sympathetic outflow during exercise. Indeed, unmyelinated nerve fibers are present in the phrenic nerve (62) and prolonged heavy intensity exercise in rats has been shown to elevate diaphragm lactic acid concentrations (and likely other metabolites) several fold above basal levels (94). In anesthetized animals, type IV phrenic afferents are activated during fatiguing diaphragm contractions (139), whereas chemical stimulation of phrenic afferents elicits increases in HR, BP, and reduces blood flow in renal and mesenteric arterial vascular beds (145). Thus, a respiratory metaboreflex could be activated during prolonged highintensity exercise in which a fatiguing diaphragm and other respiratory muscles lead to an accumulation of metabolic byproducts (53, 119). However, isolating the effect of a respiratory metaboreflex, particularly during exercise, is technically challenging in humans. A series of studies pioneered by Dempsey and colleagues utilized a novel approach in which a mechanical ventilator was used to reduce the amount of respiratory muscle work during exercise, and thus, the stimulation of the respiratory metaboreflex. Interestingly, an increase in active skeletal muscle blood flow was found when diaphragm fatigue was prevented during maximal exercise (125, 126), whereas no changes in leg blood flow and vascular resistance was seen during respiratory muscle unloading at submaximal exercise intensities (373). While the autonomic mechanisms involved in these changes remains unclear, vasoconstriction in active skeletal muscle of dogs due to stimulation of phrenic afferents with lactic acid was prevented by combined α- and β-adrenergic blockade (273). Moreover, under resting conditions, fatiguing the respiratory muscles leads to increases in muscle SNA (317) and decreases leg blood flow and vascular conductance (305,306). Thus, although more indepth investigations are needed, the respiratory metaboreflex Volume 5, April

39 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology Figure 14 Schematic illustration depicting the afferent and efferent neural responses of the arterial baroreceptors. Reductions in BP in the carotid sinus and aortic arch are sensed by the baroreceptors eliciting decreases in afferent nerve firing. This reduction in neural input to the brainstem causes an increase in sympathetic neural outflow to the heart and vasculature, while at the same time decreasing parasympathetic nerve activity to the heart. Collectively, these reflex-mediated adjustments are designed to correct the decrease in pressure sensed by the baroreceptors and bring BP back to its original value. The converse occurs when the baroreceptors are exposed to an increase in BP. Adapted, with permission, from (76). appears to mediate increases in sympathetic vasoconstrictor outflow at high-intensities of exercise, likely as a means to redirect exercising muscle blood flow to fatiguing respiratory muscles. Arterial baroreflex The arterial baroreflex plays a major role in the autonomic and ultimately BP adjustments that accompany acute cardiovascular stressors, including exercise. A number of animal and human investigations have been performed in an effort to identify the components of the baroreflex arc, establishing the basis for our current understanding of arterial baroreceptor anatomy, neural processing and function (197, 278, 304). Briefly, the carotid and aortic baroreceptors are comprised of unencapsulated free nerve endings located at the medialadventitial border of arteries in the carotid sinus bifurcation and aortic arch (278, 304) that function as the sensors in a negative feedback control system (138). Alterations in BP cause a conformational change in the baroreceptors leading to changes in afferent neuronal firing. A branch of the glossopharyngeal nerve, the Hering nerve, carries impulses from the carotid baroreceptors, while small vagal branches carry impulses from the aortic baroreceptors. These afferent signals converge centrally within the NTS of the medulla oblongata. When BP is elevated, the baroreceptors are stretched and this deformation causes an increase in afferent neuronal firing, which results in a reflex-mediated increase in parasympathetic nerve activity and decrease in SNA. Conversely, when BP is lowered, afferent firing is reduced, resulting in a decrease in parasympathetic nerve activity and an increase in SNA. In both cases, the autonomic adjustments will affect both the heart and the blood vessels altering cardiac output and vascular conductance, respectively (76, 158, 263), and returning BP to its original set point value (138) (Fig. 14). Although initially debated, the fundamental role of the arterial baroreflex to the autonomic adjustments to exercise is now well established. It has been demonstrated in both animals (40,364,365) and healthy subjects (19,76,249,258,263) that during exercise the arterial baroreflex continues to regulate BP by resetting to operate around the exercise-induced elevation in BP. Moreover, there is convincing evidence that a properly functioning arterial baroreflex is requisite for an appropriate neural cardiovascular response to exercise (54, 293, 312, 342, 365). In this regard, previous studies have reported that acute baroreceptor denervation leads to an exaggerated increase in BP in exercising dogs (51, 365, 366). Similarly, in humans who have surgically denervated carotid baroreceptors, not only is resting BP variability elevated, but the BP response to exercise is exaggerated (312, 342). An emerging concept, described in detail by Michael Joyner (158), is that the baroreflex acts to partially restrain the BP response to exercise by buffering increases in SNA produced by activation of central command and the exercise pressor reflex (14, 35, 158, 307). This concept is substantiated by the greater increase in muscle SNA and HR observed during 496 Volume 5, April 2015

40 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans handgrip in young healthy subjects when arterial baroreflex activation is prevented by pharmacologically clamping BP at resting values and negating the exercise-induced rise in BP (293). Thus, impaired arterial baroreflex function (i.e., decreased sensitivity or gain) can lead to altered neural cardiovascular responses during exercise indicating that an operational baroreflex is necessary for the appropriate autonomic and cardiovascular response to exercise. Overall, there is considerable research demonstrating that the arterial baroreflex contributes importantly to the autonomic responses to exercise and as such, plays a major role in mediating the neural cardiovascular adjustments to exercise. Arterial baroreflex at exercise onset While, much more is known about arterial baroreflex regulation under steady-state exercise conditions, a functional arterial baroreflex appears necessary for evoking the appropriate central cardiovascular and peripheral circulatory adjustments in the transition from rest to exercise (54). The magnitude of R-R interval lengthening in response to carotid baroreceptor stimulation in the decerebrate cat is attenuated by electrically evoked hindlimb contraction (213) or selective activation of group III and IV skeletal muscle afferents (214). In contrast, the magnitude of the fall in HR in response to aortic depressor nerve stimulation in decerebrate cats was similar at rest and the onset of both electrically evoked muscle contraction and muscle stretch (227). However, the bradycardic response to aortic depressor nerve stimulation was transiently attenuated at the onset of volitional isometric exercise in conscious cats, but restored later in exercise (177, 200, 227). Taken together these findings suggest that the activation of central command and/or skeletal muscle afferents can inhibit baroreflex regulation of HR via the parasympathetic nerves at the onset of exercise. Limited work has examined whether arterial baroreflex function is modified in humans during the transition from rest to exercise. We noted a transient attenuation of the magnitude of the fall in HR in response to carotid baroreceptor loading (neck suction, 60 mmhg) at the immediate onset ( 1 s)of high intensity handgrip (45%-60% MVC), but not during lowmoderate intensity handgrip (15%-30% MVC). Attribution of the blunted baroreflex responsiveness observed to either the activation of central command or the muscle mechanoreflex is not presently possible, as the activation of both would be expected to be graduated with exercise intensity. However, in an attempt to understand the mechanism involved we performed carotid baroreceptor loading (neck suction, 60 mmhg) during the anticipation of isometric handgrip exercise. Intriguingly, we noted a blunting of the magnitude of the reduction in HR evoked by neck suction when delivered immediately prior to handgrip (Fig. 15). Given the lack of muscle mechanoreflex or metaboreflex activation at this time one presumes a central mechanism is responsible, although interestingly this effect was only observed in the first two of 4 trials, indicative of a habituation response. Although, it Figure 15 Anticipation of exercise blunts carotid baroreflex mediated HR responses. Summary data showing the HR responses to neck suction (NS) at 60 Torr performed at rest and in anticipation of isometric handgrip exercise during four repeat trials. This anticipatory period was used to isolate feedforward central command input in the absence of any feedback from skeletal muscle. To create an anticipatory period, subjects were instructed that immediately after the cessation of NS they must take hold of the handgrip dynamometer and start exercising as rapidly as possible at 45% MVC and sustain the contraction for 1 min. Although a clear blunting of carotid-cardiac responses was observed in the first two trails, a habituation in the response was found. represents P < 0.05 versus rest. Unpublished observations made by the authors. is interesting to note that passive calf stretch to selectively activate the muscle mechanoreflex has also been shown to attenuate the HR response to carotid baroreceptor loading in humans ( 30 mmhg, neck suction) (110). Finally, it remains to be determined whether such modulation of arterial baroreflex responses at the onset of exercise in humans is attributable to a reduced baroreflex sensitivity (i.e., reduced maximal gain) (200, 215) or rapid resetting of the baroreflex function curve (254, 255). However, work by Dicarlo and Bishop (54) in conscious rabbits identified the importance of an immediate baroreflex resetting in mediating the increases in renal sympathetic outflow, BP, and HR at the onset of exercise. Thus, a rapid baroreflex resetting appears requisite to the initiation of the autonomic response to exercise and subsequent hemodynamic and cardiovascular adjustments. Arterial baroreflex during steady-state exercise During steady-state dynamic exercise the arterial baroreflex stimulus-response relationship is reset to function about the established BP and in general, the baroreflex maintains its ability to regulate BP as effectively as during rest (19,40,216,249,258). It was Bevegard and Shepherd (19) who initially reported that carotid baroreflex regulation was maintained during exercise in humans. These investigators demonstrated that HR, BP, and vascular conductance responses to simulated carotid sinus hypertension with the application of neck suction were similar during exercise compared to those observed at rest. Subsequently, Melcher and Donald Volume 5, April

41 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology Heart rate, muscle SNA, or mean arterial pressure Rest Ex (light) Arterial blood pressure Centering point Ex (heavy) Ex (moderate) Figure 16 A schematic summary of carotid baroreflex resetting that occurs from rest to heavy exercise. In general, the carotid baroreflex function curve for HR, muscle SNA, and MAP is progressively reset from rest to heavy exercise. However, the functional characteristics of the stimulus-response curve differ depending on the dependent variable studied. See text for details. (216) constructed full stimulus-response curves of the isolated carotid baroreceptors in chronically instrumented exercising dogs and were the first to demonstrate that the baroreflex function curve was reset by exercise to operate around the prevailing BP without a change in reflex sensitivity. Potts et al. (258) confirmed these findings in humans using the variable pressure neck chamber to simulate carotid hypotension (neck pressure) and hypertension (neck suction). These investigators demonstrated that the carotid baroreflex was reset during upright dynamic leg cycling to functionally operate around the exercising BP without a change in sensitivity. Numerous studies followed that have confirmed and extended these initial findings, demonstrating that baroreflex resetting occurs in direct relation to the intensity of exercise from rest to maximum (Fig. 16) (77,105,106,232,233,242,249). Although the variable pressure neck chamber, which selectively describes carotid baroreflex function, has been the primary method used to examine exercise resetting in humans, the assumption is made that the aortic baroreflex operates in parallel with the carotid baroreflex and therefore will similarly respond and reset with exercise (76, 263). Arterial baroreflex resetting during exercise has been demonstrated to occur due to the interactive effects of central command (106, 148, 212, 246) and sensory feedback from metabolically and mechanically sensitive skeletal muscle afferents [i.e., exercise pressor reflex; (90,105,148,212,316)]. Raven and colleagues undertook an important series of experimental studies in humans that discerned the roles for central command and the exercise pressor reflex in the exercise resetting of the baroreflex. This work, which was performed to test a hypothesis originally put forth by Rowell and O Leary, has been comprehensively outlined in several reviews (76,78,263) and so will only be briefly covered here. Importantly, although it is clear that both the carotid-cardiac and carotid-bp (i.e., vasomotor) curves are reset with exercise to operate at the prevailing BP there are some differences in baroreflex control of HR and BP that deserve discussion. Indeed, studies have demonstrated that both central command and the exercise pressor reflex are capable of inducing a bi-directional rightward and upward shift of the carotid-bp curve through a facilitative interaction (76, 78, 263). Moreover, this exercise resetting occurs without a change in maximal gain or operating point gain such that the control of BP is well maintained. In contrast, while central command is capable of relocating the carotid-cardiac baroreflex stimulus-response curve both rightward and upward, it appears that input from the exercise pressor reflex contributes to a rightward shift only. In addition, although resetting of the carotid-cardiac baroreflex function curve during exercise is accompanied by a preservation of maximal gain (i.e., sensitivity), the gain at the operating point (i.e., point about which HR is regulated) is reduced. This is associated with movement of the operating point away from the centering point (i.e., region of maximal gain) and toward the reflex threshold (243). Spontaneous cardiac baroreflex sensitivity (cbrs), calculated from the dynamic fluctuations in BP and HR (e.g., sequence technique), is associated with the operating point gain and also decreases during dynamic exercise (86, 243, 288, 289). Ogoh et al. (243), using separate cholinergic (muscarinic) and β 1 -adrenergic blockade, demonstrated that the reduction in spontaneous cbrs and operating point gain of the carotid cardiac baroreflex function curve during steady-state dynamic exercise were attributable to a reduction in cardiac parasympathetic activity. Since central command and the muscle mechanoreflex have been implicated in the withdrawal of cardiac parasympathetic activity during exercise (110,224), they may reasonably be considered as prime candidates for mediating exercise-induced changes in baroreflex sensitivity. Indeed, Iellamo et al. (148) reported that spontaneous cbrs was reduced during low intensity electrically evoked exercise under free-flow conditions, and given that central command was bypassed and muscle metaboreflex activation minimal, a role of the muscle mechanoreflex was postulated. On the other hand, Gallagher and colleagues (106) noted that partial neuromuscular blockade to enhance central command during exercise evoked a reduction in operating point gain of the carotid cardiac baroreflex function curve. Thus, it appears likely that input from both central command and muscle mechanoreceptors can contribute to the relocation of the operating point gain on the carotid cardiac baroreflex function curve during dynamic exercise and the resultant reduction in cbrs that has also been observed with spontaneous cbrs measures. In contrast, neither spontaneous cbrs (149) nor the gain at the operating point of the carotid cardiac baroreflex function curve are altered during isolated muscle metaboreflex activation with postexercise ischemia following handgrip (90). However, when the muscle metaboreflex is activated by hypoperfusing a large mass of dynamically exercising skeletal muscle in either canines (288, 289) or humans (127, 128), a reduction in spontaneous cbrs is observed. The reason for these apparently discrepant findings likely relates to differences among these protocols in the prevailing levels of 498 Volume 5, April 2015

42 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans cardiac autonomic activity and degree of input from neural control mechanisms. Indeed, during postexercise ischemia the loss of inhibitory input from central command and muscle mechanoreflex and/or baroreflex activation once exercise has ceased would lead to a relative increase in cardiac parasympathetic activity (230, 238) that could mask a muscle metaboreflex mediated reduction in cbrs as noted earlier in this review for metaboreflex-mediated HR responses. In addition, it remains a matter of debate whether the partial restriction of blood flow to the exercising muscle; (i) increases central command along with the muscle metaboreflex (179, 253), (ii) causes an attenuation of cbrs by directly reducing cardiac parasympathetic activity or via an indirect mechanism (225), and (iii) reduces the operating point gain or maximal gain of the full stimulus-response function curve. It is also possible that during exercise the accumulation of metabolites within the exercising skeletal muscles sensitizes mechanically sensitive muscle afferents that are known to modulate cardiac parasympathetic activity. However, Drew et al. (60) observed that the magnitude of the passive calf stretch mediated reduction in cbrs was no different when performed during post exercise ischemia following calf exercise at 0%, 30%, 50%, and 70% MVC to grade the concentration of metabolites within the muscle. Nevertheless, the presently available evidence implicates central command, the muscle mechanoreflex and the muscle metaboreflex in the exercise induced resetting of the carotid-cardiac baroreflex function curve and the accompanying reduction in operating point gain and spontaneous cbrs, although the relative contribution of each likely varies as a function of the exercise modality studied. In contrast to the extensive body of work examining the arterial baroreflex in terms of its sensitivity (i.e., gain), its temporal response pattern (i.e., latency) has received less attention. A delay in the latency of the peak cardiac baroreflex response has previously been reported during dynamic exercise in young individuals (193, 325), although this has not been a universal finding (256). Sundblad and Linnarsson (327) proposed that this delay resulted from an exercise-induced increase in sympathetic activation. More recently, pharmacological antagonism of cardiac parasympathetic control has been reported to prolong the latency of the peak carotid-hr response (85,167), thus raising the possibility that withdrawal of cardiac parasympathetic tone may account for the more sluggish cardiac-baroreflex responses during exercise. Studies in both animals and humans have also investigated the means by which the arterial baroreflex mediates changes in BP both at rest and during exercise. In other words, given that BP is the product of cardiac output and total peripheral resistance, how much does the arterial baroreflex rely on changes in HR and stroke volume (i.e., cardiac output) compared to total vascular resistance or conductance to modulate BP. Several studies that have examined the relative contribution of changes in cardiac output and total vascular conductance to carotid baroreflex-mediated changes in BP demonstrated that the capacity of the carotid baroreflex to regulate BP depends almost exclusively on its ability to alter total vascular conductance both at rest and during exercise (35, 241, 242). In fact, although at rest approximately 25% of the carotid baroreflex-mediated BP response could be attributed to changes in cardiac output, during exercise the ability of the carotid baroreflex to control BP was solely reliant on reflex-mediated changes in peripheral conductance (242). This critical reliance of the baroreflex on vascular changes was not significantly altered by subject posture (241). Similar findings indicating the dominance of alterations in vascular conductance contributing to carotid baroreflex-mediated changes in BP have been identified in the dog using bilateral carotid occlusion at rest and during treadmill exercise (35). Thus, the ability of the arterial baroreflex to regulate BP is critically dependent on alterations in vascular tone both at rest and during exercise. Considering that the neural stimulus from the arterial baroreflex to the vasculature is the sympathetic nervous system, an understanding of baroreflex control of sympathetic outflow is critical. Although technically challenging to directly assess sympathetic outflow in humans during physical activity due to the associated movement, the application of the microneurography technique to measure baroreflex control of muscle SNA during exercise has provided some very insightful and important information. In this regard, the functional characteristics of the baroreflex control of SNA have been shown to dynamically change throughout a given bout of exercise to allow for the effective modulation of BP (Fig. 17) (147,161). Ichinose et al. (147) reported that along with a progressive resetting of the muscle SNA-diastolic BP relationship during 3 min of handgrip, a time-dependent increase in muscle SNA baroreflex sensitivity occurred. These findings of dynamic temporal changes in the baroreflex control of muscle SNA have been extended to dynamic exercise as well however; the intensity of exercise may be of greater importance for inducing changes in muscle SNA baroreflex sensitivity (245). Indeed, in humans the carotid baroreflex control of muscle SNA appears preserved during moderate-intensity one-legged kicking and arm cycling exercise (77, 166, 245), whereas during moderate to high intensity leg cycling an increase in arterial baroreflex-muscle SNA gain has been observed (146). In general agreement, the sensitivity of the arterial baroreflex control of renal SNA has been shown to be increased during high-intensity (90% maximal HR) treadmill exercise in conscious rats (218). Collectively, these studies indicate a progressive resetting of the baroreflex control of SNA to operate around the exercise-induced elevations in BP with a maintained or increased sensitivity depending on the intensity of the exercise performed. Thus, the arterial baroreflex control of sympathetic outflow is well maintained throughout a bout of exercise. Cardiopulmonary baroreflex Mechanically sensitive receptors situated in the heart (atria, ventricles), lungs, and great veins provide feedback to medullary vasomotor centers via unmyelinated vagal afferents Volume 5, April

43 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology (A) MSNA burst incidence (bursts/100 heart beats) (B) ABR sensitivity (bursts/100 heart beats/mmhg) Rest Unloaded EX DBP (mmhg) Initial 50% EX Rest Unloaded EX Initial 50% EX Later 50% EX Operating point Later 50% EX Figure 17 Progressive resetting of the arterial baroreflex control of muscle SNA (MSNA) in the transition from rest to steady-state two arm cycling exercise. (A) Summary data showing the average operating points ( ) with the corresponding mean linear regression lines relating MSNA burst incidence and diastolic BP at rest, unloaded exercise (EX), initial 50% EX, and later 50% EX. (B) Group summary data for the slopes of the linear regression lines between MSNA burst incidence and diastolic BP (i.e., arterial baroreflex sensitivity). These findings indicate that baroreflex control of MSNA is well maintained throughout dynamic exercise in humans, progressively being reset to operate around the exercise-induced elevations in BP without any changes in reflex sensitivity. See text for further details. Reprinted, with permission, from (245). (C-fibers) in response to changes in central venous pressure and volume (197). The loading of these cardiopulmonary receptors exerts a reflex inhibition of sympathetic adrenergic activity to several vascular beds and conversely their unloading evokes a marked increase in SNA. The experimental approaches utilized to evaluate cardiopulmonary baroreceptor function have been reviewed in detail elsewhere (197) and will be discussed sparingly here. Cardiopulmonary baroreflex at exercise onset Callister et al. (30) specifically evaluated the muscle SNA (radial nerve at elbow) responses to 1 min of upright leg cycling at submaximal workloads ranging from approximately 10% to 80% of peak aerobic power (Fig. 18). Irrespective of workload, a marked reduction in muscle SNA was observed during the preparation for and at the onset of exercise, while muscle SNA did not increase until 40 to 60 s of exercise at the highest workload. While the suppression of muscle SNA in anticipation of exercise may be related to the mild cognitive effort or arousal associated with the task (31), as described in detail below, the inhibition of muscle SNA at the onset of dynamic exercise is likely linked to a loading of the cardiopulmonary baroreceptors. Cardiopulmonary baroreflex during steady-state exercise The observation that HR is typically lower during dynamic exercise performed in a supine position compared to an upright position provided an early indication that changes in central blood volume and cardiopulmonary loading may influence the cardiovascular response to steady-state exercise in humans (320). The earliest studies specifically examining the functional significance of the cardiopulmonary baroreflex on the autonomic adjustments to exercise during exercise in humans typically employed handgrip exercise. Walker et al. (1980) reported that low intensity isometric handgrip evoked a vasoconstriction in the nonexercising forearm that was threefold greater when handgrip was combined with mild lower body negative pressure (LBNP, 5 mmhg) to unload the cardiopulmonary baroreceptors (367). Further, the increase in forearm vascular resistance was significantly greater during combined LBNP and handgrip, than the algebraic addition of the response to LBNP alone plus handgrip alone. Thus it was contended that the cardiopulmonary baroreceptors tonically inhibit sympathetic vasoconstriction during isometric handgrip. However, these findings and this conclusion were disputed by subsequent investigations (290, 294, 296) that observed no interaction between either the forearm vasculature or sympathetic responses to LBNP and isometric handgrip. The reasons for these discrepant findings are unclear, but are suggested to relate in part to the level of LBNP, and thus cardiopulmonary unloading, utilized (12). While isometric handgrip has a limited effect on either preload (central venous pressure) or left-ventricular enddiastolic volume in young healthy individuals (319), dynamic exercise with a relatively large muscle mass evokes a much greater increase in left ventricle preload, stroke volume, and contractility (367). Thus, it would be reasonable to expect cardiopulmonary vagal afferents to provide a marked inhibition of SNA during dynamic exercise. Despite reports that acute or chronic cardiopulmonary denervation has no effect on the cardiovascular responses of canines running on a treadmill 500 Volume 5, April 2015

44 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans (A) Arterial pressure 150 Control Preparation Initiation Exercise 80 mmhg MSNA 5 s (B) Control Preparation Initiation Exercise Arterial pressure mmhg MSNA 5 s Figure 18 Original records showing muscle SNA (MSNA; radial nerve) and BP at baseline (control) and during the preparation and initiation of leg cycling exercise at 33 W (panel A) and 166 W (panel B). A notable reduction in MSNA is evident during the preparation and initial stages of exercise at both workloads. Reprinted, with permission, from (30). (51, 364), observations in rats (36) and rabbits (55, 236) support a role for cardiac afferents in the neural control of the peripheral circulation during exercise. In humans, Saito et al. (286) observed that muscle SNA, obtained using the microneurography technique at the median nerve (elbow), was reduced below baseline during low intensity leg cycling (20% VO 2 max), but increased at higher intensities (60% and 75% VO 2 max). A muscle pump-mediated enhancement of venous return, cardiac filling pressure, and thus cardiopulmonary baroreceptor loading, in the absence of a powerful concomitant sympatho-excitatory drive from central command or the exercise pressor reflex, may account for the fall in muscle SNA at the lower workload compared to higher intensities. In agreement, Ray et al. (266) reported that both central venous pressure increased and muscle SNA fell during dynamic knee-extension exercise when subjects were in an upright-seated position, but remained unchanged from baseline when subjects were supine, once again indicative of the importance of cardiac filling pressure and central blood volume on the sympathetic adjustments to dynamic exercise in humans. Following pioneering studies in animals examining the role of the cardiopulmonary baroreceptors during exercise (51, 364), Mack et al. (195) reported that the forearm vascular responses to LBNP were similar at rest and during leg cycling exercise in humans. Subsequently, Ogoh et al. (240) manipulated cardiac filling volumes with both LBNP and infusion of 25% human serum albumin and observed no differences between the forearm and systemic vascular responses evoked at rest and exercise. One presumes that the ability of the cardiopulmonary baroreflex to modulate muscle SNA is also unchanged from rest to exercise, but this remains to be determined. Nevertheless, the available evidence indicates that the cardiopulmonary baoreflex remains functional in exercising humans but is reset to operate around an increased central venous pressure or central blood volume (195, 240). Moreover, cardiopulmonary loading during exercise results in a diminished upward and rightward resetting of the carotid-vasomotor baroreflex function curve (244, 361). Several studies have also indicated that unloading of the cardiopulmonary baroreceptors heightens the gain of the carotid baroreflex both at rest and during exercise (20, 21, 176, 251, 257), although this has not been a universal finding (64, 330). The interaction between the cardiopulmonary baroreflex and the arterial baroreflex during exercise has recently been reviewed (78). Finally, the functional significance of the cardiopulmonary baroreceptors may become particularly important when whole-body exercise is accompanied by heat stress and/or dehydration (44, 112). Secondary to the notable increase in cutaneous perfusion and reduction in plasma volume through sweating under these conditions, central blood volume, stroke volume and cardiac output are reduced, while HR and plasma norepinephrine concentration are markedly increased (44,112,276). A resultant reduction in perfusion and oxygen delivery to the exercising skeletal muscle is likely a major factor in the development of fatigue during maximal aerobic exercise (44,112). The complex effect of thermal stress on the integrative control of the cardiovascular Volume 5, April

45 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology system during exercise has been the subject of several excellent recent reviews (44, 112). In summary, while the involvement of the cardiopulmonary baroreflex in autonomic and hemodynamic regulation is perhaps not as well understood or as widely appreciated as central command, the exercise pressor reflex and the arterial baroreflex, there is ample evidence to indicate that it plays an important functional role. Indeed, the initial inhibition of muscle SNA at the onset of dynamic leg exercise may be attributed to the increase in venous return and subsequent elevations in central venous pressure and cardiopulmonary baroreceptor load. As exercise continues, this effect would be overcome by stimulation of skeletal muscle afferents and the greater central command input during higher intensity exercise contributing to increases in muscle SNA to active and inactive skeletal muscle. Arterial chemoreflex Chemically sensitive receptors located primarily in the carotid bodies, and to a lesser extent in the aortic bodies, sense changes in the chemical composition of the bloodstream and relay afferent information to medullary regions via the carotid sinus and vagus nerve, respectively. This afferent information subsequently modulates the respiratory and autonomic systems in a homeostatic effort. The carotid body is highly vascularized and the sensing of PO 2,PCO 2, and ph is thought to primarily occur via type I (i.e., glomus) cells, which form chemical synapses with the carotid sinus nerve (137, 211). However, the neurotransmission that determines the afferent signal to the medulla is complex and includes inhibitory and excitatory inputs from numerous type I cells, chemical and electrical intercellular communication among type I and between other cell types (e.g., type II cells), pre- and postsynaptic neuromodulation, as well as efferent sympathetic and parasympathetic innervation of the carotid body (183). In addition to the classical modulators, the carotid chemoreflex also responds to a vast number of circulating chemical stimuli (183) and is influenced by blood flow and shear stress (58). Moreover, medullary chemoreceptors also modulate efferent respiratory and autonomic pathways by responding to changes in the chemical milieu of the extracellular and cerebrospinal fluid, and the sensitivity of the central and peripheral chemoreceptors is likely interdependent upon each other (313). It is well accepted that the chemoreflex is intimately involved in resting autonomic, cardiovascular and ventilatory regulation (211). However, during exercise, the influence of the chemoreflex has primarily focused on respiratory adjustments, which are reviewed in detail elsewhere (92, 211), and investigations on chemoreflex-mediated cardiovascular control are sparse. Recently, Stickland, Dempsey, and colleagues conducted a series of experiments to examine the role of the chemoreflex in modulating SNA and blood flow during exercise. Pharmacological (intravenous dopamine) or physiological (acute hyperoxia) methods to inhibit the chemoreflex increased femoral blood flow and vascular conductance during rhythmic leg exercise, but had no influence under resting conditions in healthy humans (321) (Fig. 19). In addition, a P ET,O2 Blood flow (ml min 1 ) Femoral conductance [ml min 1 (100 mmhg) 1 ] Time (s) 100% O 2 100% O 2 P ET,O2 Blood flow (ml min 1 ) Femoral conductance [ml min 1 (100 mmhg) 1 ] Time (s) Figure 19 Transient inhibition of the carotid chemoreflex with inhaled hyperoxia increases exercising leg blood flow and vascular conductance. End-tidal O2 (PET,O2), femoral artery blood flow and femoral vascular conductance during transient inhaled hyperoxia at rest (left) and during two-legged knee extension exercise (right). P < 0.05 versus baseline. Reprinted, with permission, from (321). 502 Volume 5, April 2015

46 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans reduction in muscle SNA was reported in response to transient hyperoxia during moderate intensity isometric handgrip exercise (323). These findings are in line with work in exercising dogs in which inhibition of the carotid chemoreflex, with intracarotid infusion of dopamine or hyperoxic Ringer s solution, caused hindlimb vasodilation; a response that was most evident in the exercising skeletal muscle, versus the mesenteric vasculature, and attributable to chemoreflex induced sympathetic vasoconstrictor withdrawal as the responses were abolished following α-adrenergic blockade or carotid body denervation (322). Taken together, these findings suggest that the chemoreflex restrains exercising blood flow via vasoconstrictor sympathetic outflow to skeletal muscle. However, others have shown that sustained hyperoxia (3-15 min) did not influence (298) or increased (144) the muscle SNA response to exercise and reduced exercising skeletal muscle blood flow (371). Therefore, while examination of the role of the chemoreflex during exercise is in its infancy, further investigations are warranted with consideration for the contribution of the central/peripheral chemoreflex, influence of the chemoreflex on cardiac autonomic regulation, exercise modality and temporal patterns. Overall Summary We have provided a detailed review of the current understanding of the parasympathetic and sympathetic adjustments that occur with exercise along with a discussion of the contributions of several neural reflex mechanisms in mediating these autonomic changes and the ensuing cardiac and/or vascular responses in healthy humans. A short synthesis follows, including the identification of key gaps in our understanding and suggestions for further research. The onset of exercise is accompanied by an immediate increase in HR secondary to a withdrawal of cardiac parasympathetic activity principally due to the combined actions of central command and the muscle mechanoreflex. Presently, the evidence for the contribution of cardiac SNA to the initial increase in HR at exercise onset is lacking in humans and requires (re)investigation in light of recent studies in animals. An initial reduction in cardiac baroreflex control has been noted in humans as a consequence of central command and/or muscle mechanoreflex activation. However, whether this is attributable to a decrease in arterial baroreflex sensitivity or a rapid resetting of the baroreflex function curve remains unclear. Vasoconstrictor SNA to the skeletal muscle vasculature typically decreases at the initiation of dynamic exercise with a substantial muscle mass as a consequence of the muscle pump-mediated enhancement of venous return, cardiac filling pressure and loading of the cardiopulmonary baroreceptors. However, a slight muscle mechanoreceptor mediated increase in muscle SNA may be evoked at the onset of exercise modes where increases in cardiopulmonary loading are minimal (e.g., isometric contractions of a small muscle mass). The potential contribution of the arterial chemoreflex to the initial autonomic adjustments to exercise is unclear. HR increases approximately linearly with oxygen uptake during incremental dynamic exercise. Reductions in cardiac parasympathetic activity primarily mediate the increases in HR during low intensity steady-state dynamic exercise; whereas cardiac SNA contributes more as exercise intensity increases (Fig. 20). How exercise intensity influences the precise sympatho-vagal balance is still debated. Recent work has highlighted the underappreciated contribution of cardiac parasympathetic nerve activity to ventricular control, but the implications of this for cardiac control during exercise in humans remains to be determined. During intense exercise the high level of cardiac SNA likely leads to the release of neuromodulatory cotransmitters (e.g., neuropeptide Y) that may further diminish cardiac parasympathetic control (i.e., accentuated antagonism); however, our understanding of these Heart rate (beats/min) Sympathetic activity central command + muscle metaboreceptors Vagal inhibition central command + muscle tetanoreceptors Exercise ends Increase in vagal activity [no central command no muscle tetanoreceptors] Vagal increase + Sympathetic decrease muscle metaboreceptors + circulating catecholamines 30 min Moderate exercise 30 min Recovery Figure 20 Summary of the neural control mechanisms underlying the HR response to the onset of exercise, steady-state exercise, and recovery from exercise. The contribution of changes in cardiac parasympathetic and sympathetic activity and the influence of central command and feedback from metabolically (muscle metaboreceptors) and mechanically (muscle tetanoreceptors) sensitive skeletal muscle afferents are indicated. Reprinted, with permission, from (38). Volume 5, April

47 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology processes in exercising humans is limited and requires further study. During prolonged dynamic exercise at a steady-state submaximal workload a sympathetically mediated increase in HR is evoked (cardiac drift), which is exacerbated by dehydration and high ambient temperature. Although not as well documented, a reduction in β-adrenergic receptor sensitivity has been associated with an attenuated HR increase and a reduction in aerobic exercise capacity (e.g., in normal aging), but the exploitation of this as a therapeutic target for improving exercise capacity has to date been unfruitful. Nevertheless, a normally functioning autonomic nervous system is requisite for an appropriate cardiac response during exercise, which has important implications for exercise performance. The intensity-dependent changes in cardiac autonomic activity described above facilitate an increase in HR, cardiac contractility, stroke volume, and ultimately cardiac output. Collectively the sympathetic responses to exercise, along with the metabolic modulation of sympathetic vasoconstrictor activity in the active muscle (i.e., functional sympatholysis), enhance the redistribution of cardiac output to the exercising skeletal muscles. There is compelling evidence for intensity and duration-dependent increases in muscle SNA to both the active and inactive limbs during exercise that are accompanied by increased norepinephrine spillover from the cardiac, renal, and splanchnic vasculature. However, it is important to appreciate regional differences in the temporal pattern of exercise-induced sympathetic activation and the impact that both intensity and duration have on the sympathetic response. The neural reflex mechanisms underpinning the parasympathetic and sympathetic adjustments to exercise are complex with clear evidence for the involvement of central command, the exercise pressor reflex, the arterial baroreflex, and cardiopulmonary baroreceptors, along with further potential modulation via arterial chemoreceptors and phrenic afferents from respiratory muscles (i.e., respiratory metaboreflex). As discussed in this review, these neural mechanisms are all capable of modulating the autonomic adjustments to exercise and appear to work interactively to orchestrate an appropriate neural cardiovascular response to exercise in an intensitydependent manner. Increases in sympathetic outflow to skeletal muscle vasculature are manifest with a latency of 30 to 60 s during exercise, arguably due to the time taken for the accumulation of metabolites and the activation of metabolically sensitive skeletal muscle afferents. A plethora of candidate substances have been implicated in the stimulation of skeletal muscle afferents during exercise, but the definitive resolution of the specific cocktail of substances needed for a normal response has remained elusive. A small effect of the muscle mechanoreceptors on muscle SNA has been reported, that is more substantial when the presence of metabolites is increased. At high exercise intensities a contribution from central command to the increase in sympathetic vasoconstrictor drive during exercise has also been observed. Together central command and the exercise pressor reflex elevate muscle SNA during high-intensity exercise and overcome any potential sympatho-inhibitory effects of cardiopulmonary receptor loading. A role for central command in evoking a cholinergic vasodilatation has been muted, but remains controversial and further human investigations are warranted, including considerations for exercise modality, intensity and the magnitude of the muscle mass engaged. Due to the complexity of central command, the identification of specific brain region(s) responsible for evoking autonomic adjustments has remained elusive and additional work is required to better develop an integrated neurocircuitry model. An intensity-dependent resetting of the arterial baroreflex stimulus-response relationship around the established BP occurs during steady-state exercise, meaning that the baroreflex maintains its ability to regulate BP as effectively as during rest. It is clear that both central command and the exercise pressor reflex are actively involved in the resetting of the arterial baroreflex during exercise with the cardiopulmonary baroreceptors playing a modulatory role. Anatomically, the NTS of the medullary brainstem has emerged as the primary candidate for the convergence and integration of input from each of these neural reflex mechanisms and, as such, likely plays an essential role in governing the autonomic output mediating the cardiovascular responses to exercise. Investigation of the autonomic adjustments to exercise is an ongoing area of research with more recent work attempting to understand potential dysregulation in the neural mechanisms and the ensuing autonomic adjustments to exercise that appear to accompany several cardiovascular disease states (e.g., hypertension). These studies are founded on the fundamentals of the normal regulation of the autonomic and cardiovascular responses to exercise in health outlined in detail in this review. Acknowledgements The authors express gratitude and appreciation to Dr. Peter Raven, Dr. Jere Mitchell, and Prof. John Coote for their continued support and guidance, which has been instrumental to each of our careers. The authors gratefully acknowledge the research support received from United States of America National Institutes of Health Grant #HL (P.J.F.), Grant #K99HL (C.N.Y.), British Heart Foundation PG/11/41/28893 (J.P.F.), and Arthritis Research UK # (J.P.F.). References 1. Abrahams VC, Hilton SM, Zbrozyna A. Active muscle vasodilatation produced by stimulation of the brain stem: Its significance in the defence reaction. J Physiol 154: , Adreani CM, Hill JM, Kaufman MP. Responses of group III and IV muscle afferents to dynamic exercise. J Appl Physiol (1985) 82: , Adreani CM, Kaufman MP. Effect of arterial occlusion on responses of group III and IV afferents to dynamic exercise. J Appl Physiol (1985) 84: , Alam M, Smirk FH. Observations in man on a pulse-accelerating reflex from the voluntary muscles of the legs. J Physiol 92: , Volume 5, April 2015

48 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans 5. Alam M, Smirk FH. Observations in man upon a blood pressure raising reflex arising from the voluntary muscles. J Physiol 89: , al-ani M, Robins K, al-khalidi AH, Vaile J, Townend J, Coote JH. Isometric contraction of arm flexor muscles as a method of evaluating cardiac vagal tone in man. Clin Sci (Lond) 92: , Amann M, Blain GM, Proctor LT, Sebranek JJ, Pegelow DF, Dempsey JA. Group III and IV muscle afferents contribute to ventilatory and cardiovascular response to rhythmic exercise in humans. J Appl Physiol (1985) 109: , Amann M, Runnels S, Morgan DE, Trinity JD, Fjeldstad AS, Wray DW, Reese VR, Richardson RS. On the contribution of group III and IV muscle afferents to the circulatory response to rhythmic exercise in humans. J Physiol 589: , Anderson SD, Bye PT, Perry CP, Hamor GP, Theobald G, Nyberg G. Limitation of work performance in normal adult males in the presence of beta-adrenergic blockade. Aust N Z J Med 9: , Ando M, Yamamoto T, Hino A, Sato T, Nakamura Y, Matsuzaki M. Norepinephrine spillover during exercise as a novel parameter to evaluate the severity of heart failure. J Nucl Cardiol 17: , Arai Y, Saul JP, Albrecht P, Hartley LH, Lilly LS, Cohen RJ, Colucci WS. Modulation of cardiac autonomic activity during and immediately after exercise. Am J Physiol 256: H132-H141, Arrowood JA, Mohanty PK, McNamara C, Thames MD. Cardiopulmonary reflexes do not modulate exercise pressor reflexes during isometric exercise in humans. J Appl Physiol (1985) 74: , Asmussen E, Johansen SH, Jorgensen M, Nielsen M. On the nervous factors controlling respiration and circulation during exercise. Experiments with curarization. Acta Physiol Scand 63: , Augustyniak RA, Ansorge EJ, O Leary DS. Muscle metaboreflex control of cardiac output and peripheral vasoconstriction exhibit different latencies. Am J Physiol Heart Circ Physiol 278: H , Bandler R, Carrive P. Integrated defence reaction elicited by excitatory amino acid microinjection in the midbrain periaqueductal grey region of the unrestrained cat. Brain Res 439: , Basnayake SD, Hyam JA, Pereira EA, Schweder PM, Brittain JS, Aziz TZ, Green AL, Paterson DJ. Identifying cardiovascular neurocircuitry involved in the exercise pressor reflex in humans using functional neurosurgery. J Appl Physiol (1985) 110: , Baum K, Selle K, Leyk D, Essfeld D. Comparison of blood pressure and heart rate responses to isometric exercise and passive muscle stretch in humans. Eur J Appl Physiol Occ Physiol 70: , Beaulieu P, Lambert C. Peptidic regulation of heart rate and interactions with the autonomic nervous system. Cardiovas Res 37: , Bevegard BS, Shepherd JT. Circulatory effects of stimulating the carotid arterial stretch receptors in man at rest and during exercise. J Clin Invest 45: , Bevegard S, Castenfors J, Lindblad LE. Effect of carotid sinus stimulation on cardiac output and peripheral vascular resistance during changes in blood volume distribution in man. Acta Physiol Scand 101: 50-57, Bevegard S, Castenfors J, Lindblad LE, Tranesjo J. Blood pressure and heart rate regulating capacity of the carotid sinus during changes in blood volume distribution in man. Acta Physiol Scand 99: , Billman GE, Dujardin JP. Dynamic changes in cardiac vagal tone as measured by time-series analysis. Am J Physiol 258: H , Blomqvist CG, Lewis SF, Taylor WF, Graham RM. Similarity of the hemodynamic responses to static and dynamic exercise of small muscle groups. Circ Res 48: I87-92, Bloom SR, Johnson RH, Park DM, Rennie MJ, Sulaiman WR. Differences in the metabolic and hormonal response to exercise between racing cyclists and untrained individuals. J Physiol 258: 1-18, Bolme P, Fuxe K. Adrenergic and cholinergic nerve terminals in skeletal muscle vessels. Acta Physiol Scand 78: 52-59, Bowen WP. Changes in heart-rate, blood-pressure, and duration of systole resulting from bicycling. Am J Physiol 11: 59-77, Brenner I, Shek PN, Zamecnik J, Shephard RJ. Stress hormones and the immunological responses to heat and exercise. Int J Sports Med 19: , Brubaker PH, Kitzman DW. Chronotropic incompetence: Causes, consequences, and management. Circulation 123: , Buchanan F. The physiological significance of the pulse rate. Trans Oxford Univ Scientific Club 34: , Callister R, Ng AV, Seals DR. Arm muscle sympathetic nerve activity during preparation for and initiation of leg-cycling exercise in humans. J Appl Physiol (1985) 77: , Callister R, Suwarno NO, Seals DR. Sympathetic activity is influenced by task difficulty and stress perception during mental challenge in humans. J Physiol 454: , Casadei B, Cochrane S, Johnston J, Conway J, Sleight P. Pitfalls in the interpretation of spectral analysis of the heart rate variability during exercise in humans. Acta Physiol Scand 153: , Christou DD, Seals DR. Decreased maximal heart rate with aging is related to reduced {beta}-adrenergic responsiveness but is largely explained by a reduction in intrinsic heart rate. J Appl Physiol (1985) 105: 24-29, Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. Heartrate recovery immediately after exercise as a predictor of mortality. N Engl J Med 341: , Collins HL, Augustyniak RA, Ansorge EJ, O Leary DS. Carotid baroreflex pressor responses at rest and during exercise: Cardiac output vs. regional vasoconstriction. Am J Physiol Heart Circ Physiol 280:H , Collins HL, DiCarlo SE. Cardiac afferents attenuate the muscle metaboreflex in the rat. J Appl Physiol (1985) 75: , Colucci WS, Ribeiro JP, Rocco MB, Quigg RJ, Creager MA, Marsh JD, Gauthier DF, Hartley LH. Impaired chronotropic response to exercise in patients with congestive heart failure. Role of postsynaptic betaadrenergic desensitization. Circulation 80: , Coote JH. Recovery of heart rate following intense dynamic exercise. Exp Physiol 95: , Coote JH. Myths and realities of the cardiac vagus. J Physiol 591: , Coote JH, Dodds WN. The baroreceptor reflex and the cardiovascular changes associated with sustained muscular contraction in the cat. Pflugers Arch 363: , Coote JH, Hilton SM, Perez-Gonzalez JF. The reflex nature of the pressor response to muscular exercise. J Physiol 215: , Cowley AW, Liard JF, Guyton AC. Role of the baroreceptor reflex in daily control of arterial blood pressure and other variables in dogs. Circ Res 32: , Coyle EF, Gonzalez-Alonso J. Cardiovascular drift during prolonged exercise: New perspectives. Exerc Sport Sci Rev 29: 88-92, Crandall CG, Gonzalez-Alonso J. Cardiovascular function in the heatstressed human. Acta Physiol (Oxf) 199: , Critchley HD, Corfield DR, Chandler MP, Mathias CJ, Dolan RJ. Cerebral correlates of autonomic cardiovascular arousal: A functional neuroimaging investigation in humans. J Physiol 523 (Pt 1): , Cui J, Blaha C, Moradkhan R, Gray KS, Sinoway LI. Muscle sympathetic nerve activity responses to dynamic passive muscle stretch in humans. J Physiol 576: , Cui J, Mascarenhas V, Moradkhan R, Blaha C, Sinoway LI. Effects of muscle metabolites on responses of muscle sympathetic nerve activity to mechanoreceptor(s) stimulation in healthy humans. Am J Physiol Regul Integr Comp Physiol 294: R , Cui J, Moradkhan R, Mascarenhas V, Momen A, Sinoway LI. Cyclooxygenase inhibition attenuates sympathetic responses to muscle stretch in humans. Am J Physiol Heart Circ Physiol 294: H2693-H2700, Daly WJ, Overley T. Modification of ventilatory regulation by hypnosis. J Lab Clin Med 68: , Dampney RA. Functional organization of central pathways regulating the cardiovascular system. Physiol Rev 74: , Daskalopoulos DA, Shepherd JT, Walgenbach SC. Cardiopulmonary reflexes and blood pressure in exercising sinoaortic-denervated dogs. J Appl Physiol Respir Environ Exerc Physiol 57: , Davy KP, Johnson DG, Seals DR. Cardiovascular, plasma norepinephrine, and thermal adjustments to prolonged exercise in young and older healthy humans. Clin Physiol 15: , Dempsey JA, Romer L, Rodman J, Miller J, Smith C. Consequences of exercise-induced respiratory muscle work. Respir Physiol Neurobiol 151: , DiCarlo SE, Bishop VS. Onset of exercise shifts operating point of arterial baroreflex to higher pressures. Am J Physiol 262: H303-H307, DiCarlo SE, Bishop VS. Regional vascular resistance during exercise: Role of cardiac afferents and exercise training. Am J Physiol 258: H842-H847, DiCarlo SE, Chen CY, Collins HL. Onset of exercise increases lumbar sympathetic nerve activity in rats. Med Sci Sports Exerc 28: , DiMarco AF, Romaniuk JR, Von Euler C, Yamamoto Y. Immediate changes in ventilation and respiratory pattern associated with onset and cessation of locomotion in the cat. J Physiol 343: 1-16, Ding Y, Li YL, Schultz HD. Role of blood flow in carotid body chemoreflex function in heart failure. J Physiol 589: , Doyle AE. Sir Horace Smirk. Pioneer in drug treatment of hypertension. Hypertension 17: , Drew RC, Bell MP, White MJ. Modulation of spontaneous baroreflex control of heart rate and indexes of vagal tone by passive calf muscle stretch during graded metaboreflex activation in humans. J Appl Physiol (1985) 104: , Duncan G, Johnson RH, Lambie DG. Role of sensory nerves in the cardiovascular and respiratory changes with isometric forearm exercise in man. Clin Sci (Lond) 60: , Volume 5, April

49 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology 62. Duron B. Intercostal and diaphragmatic muscle endings and afferents. In: Hornbein TF, editor. Regulation of Breathing. New York: Dekker, 1981, pp Dyke CK, Dietz NM, Lennon RL, Warner DO, Joyner MJ. Forearm blood flow responses to handgripping after local neuromuscular blockade. J Appl Physiol (1985) 84: , Eiken O, Sun JC, Mekjavic IB. Effects of blood-volume distribution on the characteristics of the carotid baroreflex in humans at rest and during exercise. Acta Physiol Scand 150: 89-94, Ekblom B, Goldbarg AN, Kilbom A, Astrand PO. Effects of atropine and propranolol on the oxygen transport system during exercise in man. Scand J Clin Lab Invest 30: 35-42, Ekelund LG. Circulatory and respiratory adaptation during prolonged exercise of moderate intensity in the sitting position. Acta Physiol Scand 69: , Eldridge FL, Millhorn DE, Kiley JP, Waldrop TG. Stimulation by central command of locomotion, respiration and circulation during exercise. Respir Physiol 59: , Eldridge FL, Millhorn DE, Waldrop TG. Exercise hyperpnea and locomotion: Parallel activation from the hypothalamus. Science 211: , Electrophysiology TFoESoCatNASoPa. Heart rate variability - standards of measurement, physiological interpretation, and clinical use. Circulation 93: , Eliasson S, Folkow B, Lindgren P, Uvnas B. Activation of sympathetic vasodilator nerves to the skeletal muscles in the cat by hypothalamic stimulation. Acta Physiol Scand 23: , Ellestad MH, Wan MK. Predictive implications of stress testing. Follow-up of 2700 subjects after maximum treadmill stress testing. Circulation 51: , Epstein S, Robinson BF, Kahler RL, Braunwald E. Effects of betaadrenergic blockade on the cardiac response to maximal and submaximal exercise in man. J Clin Invest 44: , Esler M, Jennings G, Lambert G, Meredith I, Horne M, Eisenhofer G. Overflow of catecholamine neurotransmitters to the circulation: Source, fate, and functions. Physiol Rev 70: , Esler M, Lambert G, Brunner-La Rocca HP, Vaddadi G, Kaye D. Sympathetic nerve activity and neurotransmitter release in humans: Translation from pathophysiology into clinical practice. Acta Physiol Scand 177: , Fadel PJ. Arterial baroreflex control of the peripheral vasculature in humans: Rest and exercise. Med Sci Sports Exerc 40: , Fadel PJ, Ogoh S, Keller DM, Raven PB. Recent insights into carotid baroreflex function in humans using the variable pressure neck chamber. Exp Physiol 88: , Fadel PJ, Ogoh S, Watenpaugh DE, Wasmund W, Olivencia-Yurvati A, Smith ML, Raven PB. Carotid baroreflex regulation of sympathetic nerve activity during dynamic exercise in humans. Am J Physiol Heart Circ Physiol 280: H1383-H1390., Fadel PJ, Raven PB. Human investigations into the arterial and cardiopulmonary baroreflexes during exercise. Exp Physiol 97: 39-50, Fadel PJ, Wang Z, Tuncel M, Watanabe H, Abbas A, Arbique D, Vongpatanasin W, Haley RW, Victor RG, Thomas GD. Reflex sympathetic activation during static exercise is severely impaired in patients with myophosphorylase deficiency. J Physiol 548: , Fagraeus L, Linnarsson D. Autonomic origin of heart rate fluctuations at the onset of muscular exercise. J Appl Physiol (1985) 40: , Fallentin N, Jensen BR, Bystrom S, Sjogaard G. Role of potassium in the reflex regulation of blood pressure during static exercise in man. J Physiol 451: , Fernandes A, Galbo H, Kjaer M, Mitchell JH, Secher NH, Thomas SN. Cardiovascular and ventilatory responses to dynamic exercise during epidural anaesthesia in man. J Physiol 420: , Fisher JP, Adlan AM, Shantsila A, Secher JF, Sorensen H, Secher NH. Muscle metaboreflex and autonomic regulation of heart rate in humans. J Physiol 591: , Fisher JP, Bell MPD, White MJ. Cardiovascular responses to human calf muscle stretch during varying levels of muscle metaboreflex activation. Exp Physiol 90: , Fisher JP, Kim A, Young CN, Ogoh S, Raven PB, Secher NH, Fadel PJ. Influence of ageing on carotid baroreflex peak response latency in humans. J Physiol 587: , Fisher JP, Ogoh S, Junor C, Khaja A, Northrup M, Fadel PJ. Spontaneous baroreflex measures are unable to detect age-related impairments in cardiac baroreflex function during dynamic exercise in humans. Exp Physiol 94: , Fisher JP, Seifert T, Hartwich D, Young CN, Secher NH, Fadel PJ. Autonomic control of heart rate by metabolically sensitive skeletal muscle afferents in humans. J Physiol 588: , Fisher JP, White MJ. Muscle afferent contributions to the cardiovascular response to isometric exercise. Exp Physiol 89: , Fisher JP, White MJ. The time course and direction of lower limb vascular conductance changes during voluntary and electrically evoked isometric exercise of the contralateral calf muscle in man. J Physiol 546: , Fisher JP, Young CN, Fadel PJ. Effect of muscle metaboreflex activation on carotid-cardiac baroreflex function in humans. Am J Physiol Heart Circ Physiol 294: H , Fontana GA, Pantaleo T, Bongianni F, Cresci F, Lavorini F, Guerra CT, Panuccio P. Prostaglandin synthesis blockade by ketoprofen attenuates respiratory and cardiovascular responses to static handgrip. J Appl Physiol (1985) 78: , Forster HV, Haouzi P, Dempsey JA. Control of breathing during exercise. Compr Physiol 2: , Fox SM, III, Naughton JP, Haskell WL. Physical activity and the prevention of coronary heart disease. Ann Clin Res 3: , Fregosi RF, Dempsey JA. Effects of exercise in normoxia and acute hypoxia on respiratory muscle metabolites. J Appl Physiol (1985) 60: , Freund H, Yoshimura N, Fischer JE. Effect of exercise on postoperative nitrogen balance. J Appl Physiol Respir Environ Exerc Physiol 46: , Freund PR, Hobbs SF, Rowell LB. Cardiovascular responses to muscle ischemia in man dependency on muscle mass. J Appl Physiol Respir Environ Exerc Physiol 45: , Freyschuss U. Cardiovascular adjustment to somatomotor activation. The elicitation of increments in heart rate, aortic pressure and venomotor tone with the initiation of muscle contraction. Acta Physiol Scand Suppl 342: 1-63, Freyschuss U. Elicitation of heart rate and blood pressure increase on muscle contraction. J Appl Physiol (1985) 28: , Friedman DB, Brennum J, Sztuk F, Hansen OB, Clifford PS, Bach FW, Arendt-Nielsen L, Mitchell JH, Secher NH. The effect of epidural anaesthesia with 1% lidocaine on the pressor response to dynamic exercise in man. J Physiol 470: , Fritzsche RG, Switzer TW, Hodgkinson BJ, Coyle EF. Stroke volume decline during prolonged exercise is influenced by the increase in heart rate. J Appl Physiol (1985) 86: , Fu Q, Sugiyama Y, Kamiya A, Shamsuzzaman AS, Mano T. Effects of lower body positive pressure on muscle sympathetic nerve activity. Environ Med 42: 76-78, Fu Q, Sugiyama Y, Kamiya A, Shamsuzzaman AS, Mano T. Responses of muscle sympathetic nerve activity to lower body positive pressure. Am J Physiol 275: H1254-H1259, Galbo H, Holst JJ, Christensen NJ. Glucagon and plasma catecholamine responses to graded and prolonged exercise in man. J Appl Physiol (1985) 38: 70-76, Galbo H, Kjaer M, Secher NH. Cardiovascular, ventilatory and catecholamine responses to maximal dynamic exercise in partially curarized man. J Physiol 389: , Gallagher KM, Fadel PJ, Stromstad M, Ide K, Smith SA, Querry RG, Raven PB, Secher NH. Effects of exercise pressor reflex activation on carotid baroreflex function during exercise in humans. J Physiol 533: , Gallagher KM, Fadel PJ, Stromstad M, Ide K, Smith SA, Querry RG, Raven PB, Secher NH. Effects of partial neuromuscular blockade on carotid baroreflex function during exercise in humans. J Physiol 533: , Gandevia SC, Hobbs SF. Cardiovascular responses to static exercise in man: Central and reflex contributions. J Physiol 430: , Gasser H, Meek J. A study of the the mechanisms by which muscular exercise produces accerleration of the heart. Am J Physiol 34: 48-71, Gladwell VF, Coote JH. Heart rate at the onset of muscle contraction and during passive muscle stretch in humans: A role for mechanoreceptors. J Physiol 540: , Gladwell VF, Fletcher J, Patel N, Elvidge LJ, Lloyd D, Chowdhary S, Coote JH. The influence of small fibre muscle mechanoreceptors on the cardiac vagus in humans. J Physiol 567: , Gonzalez-Alonso J, Calbet JA, Nielsen B. Muscle blood flow is reduced with dehydration during prolonged exercise in humans. J Physiol 513(Pt 3): , Gonzalez-Alonso J, Crandall CG, Johnson JM. The cardiovascular challenge of exercising in the heat. J Physiol 586: 45-53, Goodwin GM, McCloskey DI, Mitchell JH. Cardiovascular and respiratory responses to changes in central command during isometric exercise at constant muscle tension. J Physiol 226: , Grassi G. Role of the sympathetic nervous system in human hypertension. J Hypertens 16: , Grassi G, Esler M. How to assess sympathetic activity in humans. J Hypertens 17: , Green AL, Paterson DJ. Identification of neurocircuitry controlling cardiovascular function in humans using functional neurosurgery: Implications for exercise control. Exp Physiol 93: , Volume 5, April 2015

50 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans 117. Green AL, Wang S, Owen SL, Xie K, Liu X, Paterson DJ, Stein JF, Bain PG, Aziz TZ. Deep brain stimulation can regulate arterial blood pressure in awake humans. Neuroreport 16: , Green AL, Wang S, Purvis S, Owen SL, Bain PG, Stein JF, Guz A, Aziz TZ, Paterson DJ. Identifying cardiorespiratory neurocircuitry involved in central command during exercise in humans. J Physiol 578: , Guenette JA, Sheel AW. Physiological consequences of a high work of breathing during heavy exercise in humans. J Sci Med Sport 10: , Guo A, Vulchanova L, Wang J, Li X, Elde R. Immunocytochemical localization of the vanilloid receptor 1 (VR1): Relationship to neuropeptides, the P2X3 purinoceptor and IB4 binding sites. Eur J Neurosci 11: , Hajduczok G, Hade JS, Mark AL, Williams JL, Felder RB. Central command increases sympathetic nerve activity during spontaneous locomotion in cats. Circ Res 69: 66-75, Hanna RL, Hayes SG, Kaufman MP. alpha,beta-methylene ATP elicits a reflex pressor response arising from muscle in decerebrate cats. J Appl Physiol (1985) 93: , Hansen J, Sander M, Thomas GD. Metabolic modulation of sympathetic vasoconstriction in exercising skeletal muscle. Acta Physiol Scand 168: , Hansen J, Thomas GD, Jacobsen TN, Victor RG. Muscle metaboreflex triggers parallel sympathetic activation in exercising and resting human skeletal muscle. Am J Physiol 266: H2508-H2514, Harms CA, Babcock MA, McClaran SR, Pegelow DF, Nickele GA, Nelson WB, Dempsey JA. Respiratory muscle work compromises leg blood flow during maximal exercise. J Appl Physiol (1985) 82: , Harms CA, Wetter TJ, McClaran SR, Pegelow DF, Nickele GA, Nelson WB, Hanson P, Dempsey JA. Effects of respiratory muscle work on cardiac output and its distribution during maximal exercise. J Appl Physiol (1985) 85: , Hartwich D, Aldred S, Fisher JP. Influence of menstrual cycle phase on muscle metaboreflex control of cardiac baroreflex sensitivity, heart rate and blood pressure in humans. Exp Physiol 98: , Hartwich D, Dear WE, Waterfall JL, Fisher JP. Effect of muscle metaboreflex activation on spontaneous cardiac baroreflex sensitivity during exercise in humans. J Physiol 589: , Hasking GJ, Esler MD, Jennings GL, Dewar E, Lambert G. Norepinephrine spillover to plasma during steady-state supine bicycle exercise. Comparison of patients with congestive heart failure and normal subjects. Circulation 78: , Hautala A, Tulppo MP, Makikallio TH, Laukkanen R, Nissila S, Huikuri HV. Changes in cardiac autonomic regulation after prolonged maximal exercise. Clin Physiol 21: , Hayes SG, Kindig AE, Kaufman MP. Blockade of acid sensing ion channels attenuates the exercise pressor reflex in cats. J Physiol 581: , Hayes SG, Kindig AE, Kaufman MP. Comparison between the effect of static contraction and tendon stretch on the discharge of group III and IV muscle afferents. J Appl Physiol (1985) 99: , Head GA, McCarty R. Vagal and sympathetic components of the heart rate range and gain of the baroreceptor-heart rate reflex in conscious rats. J Auton Nerv Syst 21: , Heath GW, Hagberg JM, Ehsani AA, Holloszy JO. A physiological comparison of young and older endurance athletes. J Appl Physiol Respir Environ Exerc Physiol 51: , Herr MD, Imadojemu V, Kunselman AR, Sinoway LI. Characteristics of the muscle mechanoreflex during quadriceps contractions in humans. J Appl Physiol (1985) 86: , Herring N, Paterson DJ. Neuromodulators of peripheral cardiac sympatho-vagal balance. Exp Physiol 94: 46-53, Heymans C, Bouckaert J, Dautrebande L. Sinus carotidien et reflexes respiratoires. II. Influences respiratoires reflexes de l acidose, de l alcalose, de l anhydride carbonique, de l ion hydrogene et de l anoxemie: Sinus carotidiens et echanges respiratoires dans les poumons et au dela des poumons. Arch Int Pharmacodyn Ther 39: , Heymans C. Reflexogenic areas of the cardiovascular system. Perspect Biol Med 3: , Hill JM. Discharge of group IV phrenic afferent fibers increases during diaphragmatic fatigue. Brain Res 856: , Hill JM, Adreani CM, Kaufman MP. Muscle reflex stimulates sympathetic postganglionic efferents innervating triceps surae muscles of cats. Am J Physiol 271: H38-H43, Hill JM, Kaufman MP. Attenuation of reflex pressor and ventilatory responses to static muscular contraction by intrathecal opioids. J Appl Physiol (1985) 68: , Hilton SM, Marshall JM, Timms RJ. Ventral medullary relay neurones in the pathway from the defence areas of the cat and their effect on blood pressure. J Physiol 345: , Hollander AP, Bouman LN. Cardiac acceleration in man elicited by a muscle-heart reflex. J Appl Physiol 38: , Houssiere A, Najem B, Cuylits N, Cuypers S, Naeije R, van de Borne P. Hyperoxia enhances metaboreflex sensitivity during static exercise in humans. Am J Physiol Heart Circ Physiol 291: H210-H215, Hussain SN, Chatillon A, Comtois A, Roussos C, Magder S. Chemical activation of thin-fiber phrenic afferents. 2. Cardiovascular responses. J Appl Physiol (1985) 70: 77-86, Ichinose M, Saito M, Fujii N, Ogawa T, Hayashi K, Kondo N, Nishiyasu T. Modulation of the control of muscle sympathetic nerve activity during incremental leg cycling. J Physiol 586: , Ichinose M, Saito M, Kondo N, Nishiyasu T. Time-dependent modulation of arterial baroreflex control of muscle sympathetic nerve activity during isometric exercise in humans. Am J Physiol Heart Circ Physiol 290: H1419-H1426, Iellamo F, Legramante JM, Raimondi G, Peruzzi G. Baroreflex control of sinus node during dynamic exercise in humans: Effects of central command and muscle reflexes. Am J Physiol 272: H1157-H1164, Iellamo F, Pizzinelli P, Massaro M, Raimondi G, Peruzzi G, Legramante JM. Muscle metaboreflex contribution to sinus node regulation during static exercise: Insights from spectral analysis of heart rate variability. Circulation 100: 27-32, Imai K, Sato H, Hori M, Kusuoka H, Ozaki H, Yokoyama H, Takeda H, Inoue M, Kamada T. Vagally mediated heart rate recovery after exercise is accelerated in athletes but blunted in patients with chronic heart failure. J Am Coll Cardiol 24: , Innes JA, De Cort SC, Evans PJ, Guz A. Central command influences cardiorespiratory response to dynamic exercise in humans with unilateral weakness. J Physiol 448: , Ishii K, Liang N, Oue A, Hirasawa A, Sato K, Sadamoto T, Matsukawa K. Central command contributes to increased blood flow in the noncontracting muscle at the start of one-legged dynamic exercise in humans. J Appl Physiol (1985) 112: , Iwamoto GA, Mitchell JH, Mizuno M, Secher NH. Cardiovascular responses at the onset of exercise with partial neuromuscular blockade in cat and man. J Physiol 384: 39-47, Jaing W. The Intergrative Action of the Autonomic Nervous System. Neurobiology and Homestasis. Cambridge University Press, Jennings JR, van der Molen MW, Brock K, Somsen RJ. How are tonic and phasic cardiovascular changes related to central motor command? Biol Psychol 35: , Johansson JE. Uber die Einwirkung der Muskelthatigkeit auf die Athmung und die Herzthatigkeit. Scand Arch Physiol 5: 20-66, Jose AD, Stitt F, Collison D. The effects of exercise and changes in body temperature on the intrinsic heart rate in man. Am Heart J 79: , Joyner MJ. Baroreceptor function during exercise: Resetting the record. Exp Physiol 91: 27-36, Joyner MJ, Freund BJ, Jilka SM, Hetrick GA, Martinez E, Ewy GA, Wilmore JH. Effects of beta-blockade on exercise capacity of trained and untrained men: A hemodynamic comparison. J Appl Physiol (1985) 60: , Kadowaki A, Matsukawa K, Wakasugi R, Nakamoto T, Liang N. Central command does not decrease cardiac parasympathetic efferent nerve activity during spontaneous fictive motor activity in decerebrate cats. Am J Physiol Heart Circ Physiol 300: H1373-H1385, Kamiya A, Michikami D, Fu Q, Niimi Y, Iwase S, Mano T, Suzumura A. Static handgrip exercise modifies arterial baroreflex control of vascular sympathetic outflow in humans. Am J Physiol Regul Integr Comp Physiol 281: R1134-R1139, Kaufman MP. The exercise pressor reflex in animals. Exp Physiol 97: 51-58, Kaufman MP, Forster HV, Bethesda MD. Reflexes controlling circulatory, ventilatory and airway responses to exercise. In: Rowell LB, Sheperd JT, editors. Handbook of Physiology. Exercise: Regulation and Integration of Multiple Systems. Am Physiol Soc, 1996, pp Kaufman MP, Hayes SG. The exercise pressor reflex. Clin Auton Res 12: , Kaufman MP, Longhurst JC, Rybicki KJ, Wallach JH, Mitchell JH. Effects of static muscular contraction on impulse activity of groups III and IV afferents in cats. J Appl Physiol Respir Environ Exerc Physiol 55: , Keller DM, Fadel PJ, Ogoh S, Brothers RM, Hawkins M, Olivencia- Yurvati A, Raven PB. Carotid baroreflex control of leg vasculature in exercising and non-exercising skeletal muscle in humans. J Physiol 561: , Keyl C, Schneider A, Dambacher M, Bernardi L. Time delay of vagally mediated cardiac baroreflex response varies with autonomic cardiovascular control. J Appl Physiol (1985) 91: , King AB, Menon RS, Hachinski V, Cechetto DF. Human forebrain activation by visceral stimuli. JCompNeurol413: , Kingwell BA, Thompson JM, Kaye DM, McPherson GA, Jennings GL, Esler MD. Heart rate spectral analysis, cardiac norepinephrine spillover, Volume 5, April

51 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology and muscle sympathetic nerve activity during human sympathetic nervous activation and failure. Circulation 90: , Kjaer M. Epinephrine and some other hormonal responses to exercise in man: With special reference to physical training. Int J Sports Med 10: 2-15, Kjaer M. Regulation of hormonal and metabolic responses during exercise in humans. Exerc Sport Sci Rev 20: , Kjaer M, Perko G, Secher NH, Boushel R, Beyer N, Pollack S, Horn A, Fernandes A, Mohr T, Lewis SF, et al. Cardiovascular and ventilatory responses to electrically induced cycling with complete epidural anaesthesia in humans. Acta Physiol Scand 151: , Kjær M, Pott F, Mohr T, Linkis P, Tornøe P, Secher NH. Heart rate during exercise with leg vascular occlusion in spinal cord-injured humans. J Appl Physiol (1985) 86: , Kjær M, Secher NH. Neural influence on cardiovascular and endocrine responses to static exercise in humans. Sports Medicine 13: , Koba S, Yoshida T, Hayashi N. Renal sympathetic and circulatory responses to activation of the exercise pressor reflex in rats. Exp Physiol 91: , Koike H, Mark AL, Heistad DD, Schmid PG. Influence of cardiopulmonary vagal afferent activity on carotid chemoreceptor and baroreceptor reflexes in the dog. Circ Res 37: , Komine H, Matsukawa K, Tsuchimochi H, Murata J. Central command blunts the baroreflex bradycardia to aortic nerve stimulation at the onset of voluntary static exercise in cats. Am J Physiol Heart Circ Physiol 285: H516-H526, Kounalakis SN, Nassis GP, Koskolou MD, Geladas ND. The role of active muscle mass on exercise-induced cardiovascular drift. J Sports Sci Med 7: , Kozelka JW, Christy GW, Wurster RD. Ascending pathways mediating somatoautonomic reflexes in exercising dogs. J Appl Physiol (1985) 62: , Krogh A, Lindhard J. A Comparison between voluntary and electrically evoked muscular work in man. J Physiol 51: , Krogh A, Lindhard J. The regulation of respiration and circulation during the initial stages of muscular work. J Physiol 47: , Kukielka M, Seals DR, Billman GE. Cardiac vagal modulation of heart rate during prolonged submaximal exercise in animals with healed myocardial infarctions: Effects of training. Am J Physiol Heart Circ Physiol 290: H1680-H1685, Kumar P, Prabhakar NR. Peripheral chemoreceptors: Function and plasticity of the carotid body. Compr Physiol 2: , Lassen A, Mitchell JH, Reeves DR, Jr, Rogers HB, Secher NH. Cardiovascular responses to brief static contractions in man with topical nervous blockade. J Physiol 409: , Laughlin MH. Cardiovascular response to exercise. Am J Physiol 277: S244-S259, Lehmann M, Keul J, Huber G, Da Prada M. Plasma catecholamines in trained and untrained volunteers during graduated exercise. Int J Sports Med 2: , Leonard B, Mitchell JH, Mizuno M, Rube N, Saltin B, Secher NH. Partial neuromuscular blockade and cardiovascular responses to static exercise in man. J Physiol 359: , Leuenberger U, Sinoway L, Gubin S, Gaul L, Davis D, Zelis R. Effects of exercise intensity and duration on norepinephrine spillover and clearance in humans. J Appl Physiol (1985) 75: , Levy MN. Sympathetic-parasympathetic interactions in the heart. Circ Res 29: , Li J, Maile MD, Sinoway AN, Sinoway LI. Muscle pressor reflex: Potential role of vanilloid type 1 receptor and acid-sensing ion channel. J Appl Physiol (1985) 97: , Li J, Sinoway LI. ATP stimulates chemically sensitive and sensitizes mechanically sensitive afferents. Am J Physiol Heart Circ Physiol 283: H2636-H2643, Liang N, Nakamoto T, Mochizuki S, Matsukawa K. Differential contribution of central command to the cardiovascular responses during static exercise of ankle dorsal and plantar flexion in humans. J Appl Physiol (1985) 110: , Linnarsson D, Sundberg CJ, Tedner B, Haruna Y, Karemaker JM, Antonutto G, Di Prampero PE. Blood pressure and heart rate responses to sudden changes of gravity during exercise. Am J Physiol 270: H2132- H2142, Maciel BC, Gallo L, Jr, Marin Neto JA, Lima Filho EC, Martins LE. Autonomic nervous control of the heart rate during dynamic exercise in normal man. Clin Sci (Lond) 71: , Mack G, Nose H, Nadel ER. Role of cardiopulmonary baroreflexes during dynamic exercise. J Appl Physiol (1985) 65: , Maksud MG, Wiley RL, Hamilton LH, Lockhart B. Maximal VO2, ventilation, and heart rate of Olympic speed skating candidates. J Appl Physiol 29: , Mancia G, Mark A. Arterial baroreflexes in humans. In: Sheperd J, Abboud F, editors. Handbook of Physiology. Bethseda, MD: American Physiological Society, 1983, pp Mark AL, Victor RG, Nerhed C, Wallin BG. Microneurographic studies of the mechanisms of sympathetic nerve responses to static exercise in humans. Circ Res 57: , Marshall RJ, Schirger A, Shepherd JT. Blood pressure during supine exercise in idiopathic orthostatic hypotension. Circulation 24: 76-81, Matsukawa K, Komine H, Nakamoto T, Murata J. Central command blunts sensitivity of arterial baroreceptor-heart rate reflex at onset of voluntary static exercise. Am J Physiol Heart Circ Physiol 290: H200- H208, Matsukawa K, Nakamoto T, Inomoto A. Gadolinium does not blunt the cardiovascular responses at the onset of voluntary static exercise in cats: A predominant role of central command. Am J Physiol Heart Circ Physiol 292: H121-H129, Matsukawa K, Nakamoto T, Liang N. Electrical stimulation of the mesencephalic ventral tegmental area evokes skeletal muscle vasodilatation in the cat and rat. J Physiol Sci 61: , Matsukawa K, Ninomiya I. Changes in renal sympathetic nerve activity, heart rate and arterial blood pressure associated with eating in cats. J Physiol 390: , Matsukawa K, Shindo T, Shirai M, Ninomiya I. Nitric oxide mediates cat hindlimb cholinergic vasodilation induced by stimulation of posterior hypothalamus. Jpn J Physiol 43: , Matsukawa K, Wall PT, Wilson LB, Mitchell JH. Reflex stimulation of cardiac sympathetic nerve activity during static muscle contraction in cats. Am J Physiol Heart Circ Physiol 267 H821-H827, Matsukawa K, Wilson LB, Wall PT, Mitchell JH. Differential effects of clonidine on renal sympathetic nerve activity and heart rate at onset of static exercise. J Auton Nerv Syst 51: , Mazzeo RS. Catecholamine responses to acute and chronic exercise. Med Sci Sports Exerc 23: , Mazzeo RS, Rajkumar C, Jennings G, Esler M. Norepinephrine spillover at rest and during submaximal exercise in young and old subjects. J Appl Physiol (1985) 82: , McCloskey DI, Mitchell JH. Reflex cardiovascular and respiratory responses originating in exercising muscle. JPhysiol224: , McCord JL, Tsuchimochi H, Kaufman MP. Acid-sensing ion channels contribute to the metaboreceptor component of the exercise pressor reflex. Am J Physiol Heart Circ Physiol 297: H , McDonald DM. Peripheral chemoreceptors. In: Hornbein TF, editor. Regulation of Breathing. New York: Marcel Dekker Inc, 1981, pp McIlveen SA, Hayes SG, Kaufman MP. Both central command and exercise pressor reflex reset carotid sinus baroreflex. Am J Physiol Heart Circ Physiol 280: H1454-H1463, McMahon SE, McWilliam PN. Changes in R-R interval at the start of muscle contraction in the decerebrate cat. J Physiol 447: , McWilliam PN, Yang T. Inhibition of cardiac vagal component of baroreflex by group III and IV afferents. Am J Physiol 260: H730- H734, McWilliam PN, Yang T, Chen LX. Changes in the baroreceptor reflex at the start of muscle contraction in the decerebrate cat. J Physiol 436: , Melcher A, Donald DE. Maintained ability of carotid baroreflex to regulate arterial pressure during exercise. Am J Physiol 241: H838- H849, Mense S, Stahnke M. Responses in muscle afferent fibres of slow conduction velocity to contractions and ischaemia in the cat. J Physiol 342: , Miki K, Yoshimoto M, Tanimizu M. Acute shifts of baroreflex control of renal sympathetic nerve activity induced by treadmill exercise in rats. J Physiol 548: , Mitchell JH. J.B. Wolffe memorial lecture. Neural control of the circulation during exercise. Med Sci Sports Exerc 22: , Mitchell JH. Neural circulatory control during exercise: Early insights. Exp Physiol 98: , Mitchell JH, Haskell WL, Raven PB. Classification of sports. JAmColl Cardiol 24: , Mitchell JH, Kaufman MP, Iwamoto GA. The exercise pressor reflex: Its cardiovascular effects, afferent mechanisms, and central pathways. Ann Rev Physiol 45: , Mitchell JH, Reeves DR, Jr, Rogers HB, Secher NH. Epidural anaesthesia and cardiovascular responses to static exercise in man. J Physiol 417: 13-24, Mitchell JH, Reeves DR, Jr, Rogers HB, Secher NH, Victor RG. Autonomic blockade and cardiovascular responses to static exercise in partially curarized man. J Physiol 413: , Volume 5, April 2015

52 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans 225. Miyamoto T, Kawada T, Takaki H, Inagaki M, Yanagiya Y, Jin Y, Sugimachi M, Sunagawa K. High plasma norepinephrine attenuates the dynamic heart rate response to vagal stimulation. Am J Physiol Heart Circ Physiol 284: H2412-H2418, Murata J, Matsukawa K. Cardiac vagal and sympathetic efferent discharges are differentially modified by stretch of skeletal muscle. Am J Physiol Heart Circ Physiol 280: H237-H245, Murata J, Matsukawa K, Komine H, Tsuchimochi H, Nakamoto T. Central inhibition of the aortic baroreceptors-heart rate reflex at the onset of spontaneous muscle contraction. J Appl Physiol (1985) 97: , Murphy MN, Mizuno M, Mitchell JH, Smith SA. Cardiovascular regulation by skeletal muscle reflexes in health and disease. Am J Physiol Heart Circ Physiol 301: H1191-H1204, Nagle FJ, Seals DR, Hanson P. Time to fatigue during isometric exercise using different muscle masses. Int J Sports Med 9: , Nishiyasu T, Tan N, Morimoto K, Nishiyasu M, Yamaguchi Y, Murakami N. Enhancement of parasympathetic cardiac activity during activation of muscle metaboreflex in humans. J Appl Physiol (1985) 77: , Nobrega AC, Williamson JW, Friedman DB, Araujo CG, Mitchell JH. Cardiovascular responses to active and passive cycling movements. Med Sci Sports Exerc 26: , Norton KH, Boushel R, Strange S, Saltin B, Raven PB. Resetting of the carotid arterial baroreflex during dynamic exercise in humans. J Appl Physiol (1985) 87: , Norton KH, Gallagher KM, Smith SA, Querry RG, Welch-O Connor RM, Raven PB. Carotid baroreflex function during prolonged exercise. J Appl Physiol (1985) 87: , Nowak M, Holm S, Biering-Sorensen F, Secher NH, Friberg L. Central command and insular activation during attempted foot lifting in paraplegic humans. Hum Brain Mapp 25: , Nowak M, Olsen KS, Law I, Holm S, Paulson OB, Secher NH. Command-related distribution of regional cerebral blood flow during attempted handgrip. J Appl Physiol (1985) 86: , O Hagan KP, Bell LB, Mittelstadt SW, Clifford PS. Cardiac receptors modulate the renal sympathetic response to dynamic exercise in rabbits. J Appl Physiol (1985) 76: , O Hagan KP, Bell LB, Mittelstadt SW, Clifford PS. Effect of dynamic exercise on renal sympathetic nerve activity in conscious rabbits. J Appl Physiol (1985) 74: , O Leary DS. Autonomic mechanisms of muscle metaboreflex control of heart rate. J Appl Physiol (1985) 74: , O Leary DS, Rossi NF, Churchill PC. Substantial cardiac parasympathetic activity exists during heavy dynamic exercise in dogs. Am J Physiol 273: H , Ogoh S, Brothers RM, Barnes Q, Eubank WL, Hawkins MN, Purkayastha S, O-Yurvati A, Raven PB. Cardiopulmonary baroreflex is reset during dynamic exercise. J Appl Physiol (1985) 100:51-59, Ogoh S, Fadel PJ, Monteiro F, Wasmund WL, Raven PB. Haemodynamic changes during neck pressure and suction in seated and supine positions. J Physiol 540: , Ogoh S, Fadel PJ, Nissen P, Jans O, Selmer C, Secher NH, Raven PB. Baroreflex-mediated changes in cardiac output and vascular conductance in response to alterations in carotid sinus pressure during exercise in humans. J Physiol 550: , Ogoh S, Fisher JP, Dawson EA, White MJ, Secher NH, Raven PB. Autonomic nervous system influence on arterial baroreflex control of heart rate during exercise in humans. J Physiol 566: , Ogoh S, Fisher JP, Fadel PJ, Raven PB. Increases in central blood volume modulate carotid baroreflex resetting during dynamic exercise in humans. J Physiol 581: , Ogoh S, Fisher JP, Raven PB, Fadel PJ. Arterial baroreflex control of muscle sympathetic nerve activity in the transition from rest to steadystate dynamic exercise in humans. Am J Physiol Heart Circ Physiol 293: H , Ogoh S, Wasmund WL, Keller DM, O-Yurvati A, Gallagher KM, Mitchell JH, Raven PB. Role of central command in carotid baroreflex resetting in humans during static exercise. J Physiol 543: , Olson TP, Joyner MJ, Eisenach JH, Curry TB, Johnson BD. Influence of locomotor muscle afferent inhibition on the ventilatory response to exercise in heart failure. Exp Physiol 99: , Pagani M, Montano N, Porta A, Malliani A, Abboud FM, Birkett C, Somers VK. Relationship between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation 95: , Papelier Y, Escourrou P, Gauthier JP, Rowell LB. Carotid baroreflex control of blood pressure and heart rate in men during dynamic exercise. J Appl Physiol (1985) 77: , Pawelczyk JA, Pawelczyk RA, Warberg J, Mitchell JH, Secher NH. Cardiovascular and catecholamine responses to static exercise in partially curarized humans. Acta Physiol Scand 160: 23-28, Pawelczyk JA, Raven PB. Reductions in central venous pressure improve carotid baroreflex responses in conscious men. Am J Physiol 257: H1389-H1395, Pollak KA, Swenson JD, Vanhaitsma TA, Hughen RW, Jo D, Light KC, Schweinhardt P, Amann M, Light AR. Exogenously applied muscle metabolites synergistically evoke sensations of muscle fatigue and pain in human subjects. Exp Physiol 99: , Pomeroy G, Ardell JL, Wurster RD. Spinal opiate modulation of cardiovascular reflexes in the exercising dog. Brain Res 381: , Potts JT. Inhibitory neurotransmission in the nucleus tractus solitarii: Implications for baroreflex resetting during exercise. Exp Physiol 91: 59-72, Potts JT, Mitchell JH. Rapid resetting of carotid baroreceptor reflex by afferent input from skeletal muscle receptors. Am J Physiol 275: H2000-H2008, Potts JT, Raven PB. Effect of dynamic exercise on human carotidcardiac baroreflex latency. Am J Physiol 268: H1208-H1214, Potts JT, Shi X, Raven PB. Cardiopulmonary baroreceptors modulate carotid baroreflex control of heart rate during dynamic exercise in humans. Am J Physiol 268: H1567-H1576, Potts JT, Shi XR, Raven PB. Carotid baroreflex responsiveness during dynamic exercise in humans. Am J Physiol 265: H1928-H1938, Proakis AG, Harris GB. Comparative penetration of glycopyrrolate and atropine across the blood brain and placental barriers in anesthetized dogs. Anesthesiology 48: , Pryor SL, Lewis SF, Haller RG, Bertocci LA, Victor RG. Impairment of sympathetic activation during static exercise in patients with muscle phosphorylase deficiency (McArdle s disease). J Clin Invest 85: , Querry RG, Smith SA, Stromstad M, Ide K, Raven PB, Secher NH. Neural blockade during exercise augments central command s contribution to carotid baroreflex resetting. Am J Physiol Heart Circ Physiol 280: H1635-H1644, Raczkowska M, Eckberg DL, Ebert TJ. Muscarinic cholinergic receptors modulate vagal cardiac responses in man. J Auton Nerv Syst 7: , Raven PB, Fadel PJ, Ogoh S. Arterial baroreflex resetting during exercise: A current perspective. Exp Physiol 91: 37-49, Raven PB, Fadel PJ, Smith SA. The influence of central command on baroreflex resetting during exercise. Exerc Sport Sci Rev 30: 39-44, Raven PB, Potts JT, Shi X. Baroreflex regulation of blood pressure during dynamic exercise. Exerc Sport Sci Rev 25: , Ray CA, Rea RF, Clary MP, Mark AL. Muscle sympathetic nerve responses to dynamic one-legged exercise: Effect of body posture. Am J Physiol 264: H1-H7, Ray CA, Wilson TE. Comparison of skin sympathetic nerve responses to isometric arm and leg exercise. J Appl Physiol (1985) 97: , Remensnyder JP, Mitchell JH, Sarnoff SJ. Functional sympatholysis during muscular activity. Circ Res 11: , Reybrouck T, Amery A, Billiet L. Hemodynamic response to graded exercise after chronic beta-adrenergic blockade. J Appl Physiol Respir Environ Exerc Physiol 42: , Robergs RA, Landweher R. The surprising history of the HRmax=220-age equation. J Exerc Physiol online 5: 1-10, Robinson BF, Epstein SE, Beiser GD, Braunwald E. Control of heart rate by the autonomic nervous system. Studies in man on the interrelation between baroreceptor mechanisms and exercise. Circ Res 19: , Robinson S, Pearcy M, Brueckman FR, Nicholas JR, Miller DI. Effects of atropine on heart rate and oxygen intake in working man. JAppl Physiol (1985) 5: , Rodman JR, Henderson KS, Smith CA, Dempsey JA. Cardiovascular effects of the respiratory muscle metaboreflexes in dogs: Rest and exercise. J Appl Physiol (1985) 95: , Rotto DM, Kaufman MP. Effect of metabolic products of muscular contraction on discharge of group III and IV afferents. J Appl Physiol (1985) 64: , Rotto DM, Schultz HD, Longhurst JC, Kaufman MP. Sensitization of group III muscle afferents to static contraction by arachidonic acid. J Appl Physiol (1985) 68: , Rowell LB, Marx HJ, Bruce RA, Conn RD, Kusumi F. Reductions in cardiac output, central blood volume, and stroke volume with thermal stress in normal men during exercise. J Clin Invest 45: , Sadamoto T, Matsukawa K. Cardiovascular responses during spontaneous overground locomotion in freely moving decerebrate cats. J Appl Physiol (1985) 83: , Sagawa K. Baroreflex control of systemic arterial pressure and vascular bed. In: Handbook of Physiology. The Cardiovascular System. Bethesda, MD: Am. Physiol. Soc., 1983, pp Volume 5, April

53 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology 279. Saito M, Iwase S, Mano T. Different responses of muscle sympathetic nerve activity to sustained and rhythmic handgrip exercises. Jpn J Physiol 36: , Saito M, Mano T. Exercise mode affects muscle sympathetic nerve responsiveness. Jpn J Physiol 41: , Saito M, Mano T, Iwase S. Sympathetic nerve activity related to local fatigue sensation during static contraction. J Appl Physiol (1985) 67: , Saito M, Mano T, Iwase S. Changes in muscle sympathetic nerve activity and calf blood flow during static handgrip exercise. Eur J Appl Physiol Occup Physiol 60: , Saito M, Naito M, Mano T. Different responses in skin and muscle sympathetic nerve activity to static muscle contraction. J Appl Physiol (1985) 69: , Saito M, Nakamura Y. Cardiac autonomic control and muscle sympathetic nerve activity during dynamic exercise. Jpn J Physiol 45: , Saito M, Sone R, Ikeda M, Mano T. Sympathetic outflow to the skeletal muscle in humans increases during prolonged light exercise. J Appl Physiol (1985) 82: , Saito M, Tsukanaka A, Yanagihara D, Mano T. Muscle sympathetic nerve responses to graded leg cycling. J Appl Physiol (1985) 75: , Saitoh H, Nomura A, Osaka M, Sasabe N, Atarashi H, Hayakawa H. Developmental changes of alpha 1-adrenergic chronotropic action on human sinus node. Am J Cardiol 76: 89-91, Sala-Mercado JA, Ichinose M, Coutsos M, Li Z, Fano D, Ichinose T, Dawe EJ, O Leary DS. Progressive muscle metaboreflex activation gradually decreases spontaneous heart rate baroreflex sensitivity during dynamic exercise. Am J Physiol Heart Circ Physiol 298: H594-H600, Sala-Mercado JA, Ichinose M, Hammond RL, Ichinose T, Pallante M, Stephenson LW, O Leary DS, Iellamo F. Muscle metaboreflex attenuates spontaneous heart rate baroreflex sensitivity during dynamic exercise. Am J Physiol Heart Circ Physiol 292: H2867-H2873, Sanders JS, Ferguson DW. Cardiopulmonary baroreflexes fail to modulate sympathetic responses during isometric exercise in humans: Direct evidence from microneurographic studies. J Am Coll Cardiol 12: , Savard G, Strange S, Kiens B, Richter EA, Christensen NJ, Saltin B. Noradrenaline spillover during exercise in active versus resting skeletal muscle in man. Acta Physiol Scand 131: , Savin WM, Davidson DM, Haskell WL. Autonomic contribution to heart rate recovery from exercise in humans. J Appl Physiol Respir Environ Exerc Physiol 53: , Scherrer U, Pryor SL, Bertocci LA, Victor RG. Arterial baroreflex buffering of sympathetic activation during exercise-induced elevations in arterial pressure. J Clin Invest 86: , Scherrer U, Vissing SF, Victor RG. Effects of lower-body negative pressure on sympathetic nerve responses to static exercise in humans. Microneurographic evidence against cardiac baroreflex modulation of the exercise pressor reflex. Circulation 78: 49-59, Schultz HD, Li YL, Ding Y. Arterial chemoreceptors and sympathetic nerve activity: Implications for hypertension and heart failure. Hypertension 50: 6-13, Seals DR. Cardiopulmonary baroreflexes do not modulate exerciseinduced sympathoexcitation. J Appl Physiol (1985) 64: , Seals DR. Influence of muscle mass on sympathetic neural activation during isometric exercise. J Appl Physiol (1985) 67: , Seals DR, Johnson DG, Fregosi RF. Hyperoxia lowers sympathetic activity at rest but not during exercise in humans. Am J Physiol 260: R873-R878, Seals DR, Victor RG. Regulation of muscle sympathetic nerve activity during exercise in humans. Exerc Sport Sci Rev 19: , Seals DR, Victor RG, Mark AL. Plasma norepinephrine and muscle sympathetic discharge during rhythmic exercise in humans. J Appl Physiol (1985) 65: , Secher NH. Heart rate at the onset of static exercise in man with partial neuromuscular blockade. J Physiol 368: , Secher NH, Amann M. Human investigations into the exercise pressor reflex. Exp Physiol 97: 59-69, Secher NH, Kjaer M, Galbo H. Arterial blood pressure at the onset of dynamic exercise in partially curarized man. Acta Physiol Scand 133: , Sheehan D, Mulholland JH, Safiroff B. Surgical anatomy of the carotid sinus nerve. Anat Rec 8o: , Sheel AW, Derchak PA, Morgan BJ, Pegelow DF, Jacques AJ, Dempsey JA. Fatiguing inspiratory muscle work causes reflex reduction in resting leg blood flow in humans. J Physiol 537: , Sheel AW, Derchak PA, Pegelow DF, Dempsey JA. Threshold effects of respiratory muscle work on limb vascular resistance. Am J Physiol Heart Circ Physiol 282: H1732-H1738, Sheriff DD, O Leary DS, Scher AM, Rowell LB. Baroreflex attenuates pressor response to graded muscle ischemia in exercising dogs. Am J Physiol 258: H , Shoemaker JK, Halliwill JR, Hughson RL, Joyner MJ. Contributions of acetylcholine and nitric oxide to forearm blood flow at exercise onset and recovery. Am J Physiol 273: H2388-H2395, Silman I, Sussman JL. Acetylcholinesterase: classical and nonclassical functions and pharmacology. Curr Opin Pharmacol 5: , Simpson JB. The circumventricular organs and the central actions of angiotensin. Neuroendocrinology 32: , Sinoway LI, Smith MB, Enders B, Leuenberger U, Dzwonczyk T, Gray K, Whisler S, Moore RL. Role of diprotonated phosphate in evoking muscle reflex responses in cats and humans. Am J Physiol 267: H770- H778, Smit AJ, Timmers HM, Wieling W, Wagenaar M, Marres HM, Lenders JM, Montfrans GAv, Karemaker JM. Long-term effects of carotid sinus denervation on arterial blood pressure in humans. Circulation 105: , Smith CA, Forster HV, Blain GM, Dempsey JA. An interdependent model of central/peripheral chemoreception: Evidence and implications for ventilatory control. Respir Physiol Neurobiol 173: , Smith OA, Jr, Rushmer RF, Lasher EP. Similarity of cardiovascular responses to exercise and to diencephalic stimulation. Am J Physiol 198: , Smith SA, Mitchell JH, Garry MG. The mammalian exercise pressor reflex in health and disease. Exp Physiol 91: , Smith SA, Querry RG, Fadel PJ, Gallagher KM, Stromstad M, Ide K, Raven PB, Secher NH. Partial blockade of skeletal muscle somatosensory afferents attenuates baroreflex resetting during exercise in humans. J Physiol 551: , St Croix CM, Morgan BJ, Wetter TJ, Dempsey JA. Fatiguing inspiratory muscle work causes reflex sympathetic activation in humans. J Physiol 529(Pt 2): , Stebbins CL, Brown B, Levin D, Longhurst JC. Reflex effect of skeletal muscle mechanoreceptor stimulation on the cardiovascular system. J Appl Physiol (1985) 65: , Stefadouros MA, Grossman W, el-shahawy M, Witham C. The effect of isometric exercise on the left ventricular volume in normal man. Circulation 49: , Stenberg J, Astrand PO, Ekblom B, Royce J, Saltin B. Hemodynamic response to work with different muscle groups, sitting and supine. J Appl Physiol (1985) 22: 61-70, Stickland MK, Fuhr DP, Haykowsky MJ, Jones KE, Paterson DI, Ezekowitz JA, McMurtry MS. Carotid chemoreceptor modulation of blood flow during exercise in healthy humans. JPhysiol589: , Stickland MK, Miller JD, Smith CA, Dempsey JA. Carotid chemoreceptor modulation of regional blood flow distribution during exercise in health and chronic heart failure. Circ Res 100: , Stickland MK, Morgan BJ, Dempsey JA. Carotid chemoreceptor modulation of sympathetic vasoconstrictor outflow during exercise in healthy humans. J Physiol 586: , Strack AM, Sawyer WB, Hughes JH, Platt KB, Loewy AD. A general pattern of CNS innervation of the sympathetic outflow demonstrated by transneuronal pseudorabies viral infections. Brain Res 491: , Strange S, Rowell LB, Christensen NJ, Saltin B. Cardiovascular responses to carotid sinus baroreceptor stimulation during moderate to severe exercise in man. Acta Physiol Scand 138: , Strange S, Secher NH, Pawelczyk JA, Karpakka J, Christensen NJ, Mitchell JH, Saltin B. Neural control of cardiovascular responses and of ventilation during dynamic exercise in man. J Physiol 470: , Sundblad P, Linnarsson D. Slowing of carotid-cardiac baroreflex with standing and with isometric and dynamic muscle activity. Am J Physiol 271: H1363-H1369, Takahashi M, Matsukawa K, Nakamoto T, Tsuchimochi H, Sakaguchi A, Kawaguchi K, Onari K. Control of heart rate variability by cardiac parasympathetic nerve activity during voluntary static exercise in humans with tetraplegia. J Appl Physiol (1985) 103: , Takahashi M, Sakaguchi A, Matsukawa K, Komine H, Kawaguchi K, Onari K. Cardiovascular control during voluntary static exercise in humans with tetraplegia. J Appl Physiol (1985) 97: , Takeshita A, Mark AL, Eckberg DL, Abboud FM. Effect of central venous pressure on arterial baroreflex control of heart rate. Am J Physiol 236: H42-H47, Tallarida G, Baldoni F, Peruzzi G, Brindisi F, Raimondi G, Sangiorgi M. Cardiovascular and respiratory chemoreflexes from the hindlimb sensory receptors evoked by intra-arterial injection of bradykinin and other chemical agents in the rabbit. J Pharmacol Exp Ther 208: , Volume 5, April 2015

54 Comprehensive Physiology Autonomic Adjustments to Exercise in Humans 332. Tan E, Dampney RA. Cardiovascular effects of stimulation of neurones within the defence area of the hypothalamus and midbrain of the rabbit. Clin Exp Pharmacol Physiol 10: , Tanaka H, Monahan KD, Seals DR. Age-predicted maximal heart rate revisited. J Am Coll Cardiol 37: , Taylor JA, Hand GA, Johnson DG, Seals DR. Augmented forearm vasoconstriction during dynamic exercise in healthy older men. Circulation 86: , Taylor JA, Williams TD, Seals DR, Davy KP. Low-frequency arterial pressure fluctuations do not reflect sympathetic outflow: Gender and age differences. Am J Physiol 274: H1194-H1201, Tesch PA, Kaiser P. Effect of beta-adrenergic blockade on maximal oxygen uptake in trained males. Acta Physiol Scand 112: , Thomas GD, Segal SS. Neural control of muscle blood flow during exercise. J Appl Physiol (1985) 97: , Thornton JM, Aziz T, Schlugman D, Paterson DJ. Electrical stimulation of the midbrain increases heart rate and arterial blood pressure in awake humans. J Physiol 539: , Thornton JM, Guz A, Murphy K, Griffith AR, Pedersen DL, Kardos A, Leff A, Adams L, Casadei B, Paterson DJ. Identification of higher brain centres that may encode the cardiorespiratory response to exercise in humans. J Physiol 533: , Thrasher TN. Arterial baroreceptor input contributes to long-term control of blood pressure. Curr Hypertens Rep 8: , Tidgren B, Hjemdahl P, Theodorsson E, Nussberger J. Renal neurohormonal and vascular responses to dynamic exercise in humans. J Appl Physiol (1985) 70: , Timmers HJ, Wieling W, Karemaker JM, Lenders JW. Cardiovascular responses to stress after carotid baroreceptor denervation in humans. Ann N Y Acad Sci 1018: , Tipton CM. The autonomic nervous system. In: Tipton CM, editor. Exercise Physiology: People and Ideas. Oxford University Press, 2003, pp Toma K, Walkowski S, Metzler-Wilson K, Wilson TE. Acupuncture attenuates exercise-induced increases in skin sympathetic nerve activity. Auton Neurosci 162: 84-88, Tsuchimochi H, Hayes SG, McCord JL, Kaufman MP. Both central command and exercise pressor reflex activate cardiac sympathetic nerve activity in decerebrate cats. Am J Physiol Heart Circ Physiol 296: H1157-H1163, Tsuchimochi H, Matsukawa K, Komine H, Murata J. Direct measurement of cardiac sympathetic efferent nerve activity during dynamic exercise. Am J Physiol Heart Circ Physiol 283: H , Tsuji H, Larson MG, Venditti FJ, Jr, Manders ES, Evans JC, Feldman CL, Levy D. Impact of reduced heart rate variability on risk for cardiac events. The Framingham Heart Study. Circulation 94: , Uvnas B. Cholinergic vasodilator nerves. Fed Proc 25: , Vallbo AB, Hagbarth KE, Torebjork HE, Wallin BG. Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves. Physiol Rev 59: , Victor RG, Bertocci LA, Pryor SL, Nunnally RL. Sympathetic nerve discharge is coupled to muscle cell ph during exercise in humans. J Clin Invest 82: , Victor RG, Pryor SL, Secher NH, Mitchell JH. Effects of partial neuromuscular blockade on sympathetic nerve responses to static exercise in humans. Circ Res 65: , Victor RG, Rotto DM, Pryor SL, Kaufman MP. Stimulation of renal sympathetic activity by static contraction: Evidence for mechanoreceptor-induced reflexes from skeletal muscle. Circ Res 64: , Victor RG, Seals DR. Reflex stimulation of sympathetic outflow during rhythmic exercise in humans. Am J Physiol 257: H2017-H2024, Victor RG, Seals DR, Mark AL. Differential control of heart rate and sympathetic nerve activity during dynamic exercise. Insight from intraneural recordings in humans. J Clin Invest 79: , Victor RG, Secher NH, Lyson T, Mitchell JH. Central command increases muscle sympathetic nerve activity during intense intermittent isometric exercise in humans. Circ Res 76: , Vissing J, MacLean DA, Vissing SF, Sander M, Saltin B, Haller RG. The exercise metaboreflex is maintained in the absence of muscle acidosis: Insights from muscle microdialysis in humans with McArdle s disease. J Physiol 537: , Vissing J, Vissing SF, MacLean DA, Saltin B, Quistorff B, Haller RG. Sympathetic activation in exercise is not dependent on muscle acidosis. Direct evidence from studies in metabolic myopathies. J Clin Invest 101: , Vissing SF, Hjortso EM. Central motor command activates sympathetic outflow to the cutaneous circulation in humans. J Physiol 492 (Pt 3): , Vissing SF, Scherrer U, Victor RG. Stimulation of skin sympathetic nerve discharge by central command. Differential control of sympathetic outflow to skin and skeletal muscle during static exercise. Circ Res 69: , Vivekananthan DP, Blackstone EH, Pothier CE, Lauer MS. Heart rate recovery after exercise is a predictor of mortality, independent of the angiographic severity of coronary disease. J Am Coll Cardiol 42: , Volianitis S, Yoshiga CC, Vogelsang T, Secher NH. Arterial blood pressure and carotid baroreflex function during arm and combined arm and leg exercise in humans. Acta Physiol Scand 181: , Waldrop TG, Eldridge FL, Iwamoto GA, Mitchell JH. Central neural control of respiration and circulation during exercise. In: Rowell LB, Sheperd JT, editors. Handbook of Physiology. Exercise: Regulation and Integration of Multiple Systems. Bethesda: Am Physiol Soc, 1996, pp Waldrop TG, Mullins DC, Millhorn DE. Control of respiration by the hypothalamus and by feedback from contracting muscles in cats. Respir Physiol 64: , Walgenbach SC, Donald DE. Cardiopulmonary reflexes and arterial pressure during rest and exercise in dogs. Am J Physiol 244: H , Walgenbach SC, Donald DE. Inhibition by carotid baroreflex of exercise-induced increases in arterial pressure. Circ Res 52: , Walgenbach SC, Shepherd JT. Role of arterial and cardiopulmonary mechanoreceptors in the regulation of arterial pressure during rest and exercise in conscious dogs. Mayo Clin Proc 59: , Walker JL, Abboud FM, Mark AL, Thames MD. Interaction of cardiopulmonary and somatic reflexes in humans. J Clin Invest 65: , Wallin BG, Esler M, Dorward P, Eisenhofer G, Ferrier C, Westerman R, Jennings G. Simultaneous measurements of cardiac noradrenaline spillover and sympathetic outflow to skeletal muscle in humans. JPhysiol 453: 45-58, Wallin BG, Thompson JM, Jennings GL, Esler MD. Renal noradrenaline spillover correlates with muscle sympathetic activity in humans. J Physiol 491(Pt 3): , Wang Y, Morgan WP. The effect of imagery perspectives on the psychophysiological responses to imagined exercise. Behav Brain Res 52: , Welch HG, Bonde-Petersen F, Graham T, Klausen K, Secher N. Effects of hyperoxia on leg blood flow and metabolism during exercise. J Appl Physiol Respir Environ Exerc Physiol 42: , Welsh DG, Segal SS. Coactivation of resistance vessels and muscle fibers with acetylcholine release from motor nerves. Am J Physiol 273: H156-H163, Wetter TJ, Harms CA, Nelson WB, Pegelow DF, Dempsey JA. Influence of respiratory muscle work on VO(2) and leg blood flow during submaximal exercise. J Appl Physiol (1985) 87: , White DW, Raven PB. Autonomic neural control of heart rate during dynamic exercise: Revisited. J Physiol 592: , Wildenthal K, Mierzwiak DS, Skinner NS, Jr, Mitchell JH. Potassiuminduced cardiovascular and ventilatory reflexes from the dog hindlimb. Am J Physiol 215: , Williams MA, Smith SA, O Brien DE, Mitchell JH, Garry MG. The group IV afferent neuron expresses multiple receptor alterations in cardiomyopathyic rats: Evidence at the cannabinoid CB1 receptor. JPhysiol 586: , Williamson JW. The relevance of central command for the neural cardiovascular control of exercise. Exp Physiol 95: , Williamson JW, Fadel PJ, Mitchell JH. New insights into central cardiovascular control during exercise in humans: A central command update. Exp Physiol 91: 51-58, Williamson JW, McColl R, Mathews D, Mitchell JH, Raven PB, Morgan WP. Brain activation by central command during actual and imagined handgrip under hypnosis. J Appl Physiol (1985) 92: , Williamson JW, McColl R, Mathews D, Mitchell JH, Raven PB, Morgan WP. Hypnotic manipulation of effort sense during dynamic exercise: Cardiovascular responses and brain activation. J Appl Physiol (1985) 90: , Williamson JW, Mitchell JH, Olesen HL, Raven PB, Secher NH. Reflex increase in blood pressure induced by leg compression in man. J Physiol 475: , Williamson JW, Nobrega AC, Winchester PK, Zim S, Mitchell JH. Instantaneous heart rate increase with dynamic exercise: Central command and muscle-heart reflex contributions. J Appl Physiol (1985) 78: , Williamson JW, Olesen HL, Pott F, Mitchell JH, Secher NH. Central command increases cardiac output during static exercise in humans. Acta Physiol Scand 156: , Volume 5, April

55 Autonomic Adjustments to Exercise in Humans Comprehensive Physiology 384. Wilson TE, Dyckman DJ, Ray CA. Determinants of skin sympathetic nerve responses to isometric exercise. J Appl Physiol (1985) 100: , Winchester PK, Williamson JW, Mitchell JH. Cardiovascular responses to static exercise in patients with Brown-Sequard syndrome. J Physiol 527(Pt 1): , Yaksh TL, Noueihed R. The physiology and pharmacology of spinal opiates. Annu Rev Pharmacol Toxicol 25: , Yamada T, Shimonagata T, Fukunami M, Kumagai K, Ogita H, Hirata A, Asai M, Makino N, Kioka H, Kusuoka H, Hori M, Hoki N. Comparison of the prognostic value of cardiac iodine-123 metaiodobenzylguanidine imaging and heart rate variability in patients with chronic heart failure: A prospective study. J Am Coll Cardiol 41: , Yoh M, Yuasa F, Mimura J, Yokoe H, Kawamura A, Sugiura T, Iwasaka T. Resting muscle sympathetic nerve activity, cardiac metaiodobenzylguanidine uptake, and exercise tolerance in patients with left ventricular dysfunction. J Nucl Cardiol 16: , Zouhal H, Jacob C, Delamarche P, Gratas-Delamarche A. Catecholamines and the effects of exercise, training and gender. Sports Med 38: , Zuntz N, Geppert J. Ueber die natur der normalen atemreize und den ort ihrer wirkung. Arch Gen Physiol 38: , Volume 5, April 2015

56 Autonomic Neuroscience: Basic and Clinical 188 (2015) 3 4 Contents lists available at ScienceDirect Autonomic Neuroscience: Basic and Clinical journal homepage: Autonomic responses to exercise: Where is central command? J.W. Williamson Department of Health Care Sciences, School of Health Professions, University of Texas Southwestern Medical Center at Dallas, United States article info abstract Article history: Received 19 May 2014 Received in revised form 18 August 2014 Accepted 13 October 2014 Keywords: Central command Exercise A central command is thought to involve a signal arising in a central area of the brain eliciting a parallel activation of the autonomic nervous system and skeletal muscle contraction during exercise. Although much of the neural circuitry involved in autonomic control has been identified, defining the specific higher brain region(s) serving in a central command capacity has proven more challenging. Investigators have been faced with redundancies in regulatory systems, feedback mechanisms and the complexities ofhuman neural connectivity. Several studies have attempted to address these issues and provide more definitive neuroanatomical information. However, none have clearly answered the question, where is central command? 2014 Elsevier B.V. All rights reserved. A century has passed since the concept of central command was initially proposed, yet we have struggled to clearly identify the neural substrate underlying this neurophysiological phenomenon. While significant advances in the understanding of cortical circuitry associated with autonomic function and cardiovascular control have been made, and more recently summarized in a review by Shoemaker et al. (2012), the cortical site(s) and/or networks subserving central command per se remain unclear. As typically described, central command involves descending neural signals from higher brain centers, originally defined as cortical irradiation, capable of influencing cardiovascular responses during exercise (Krogh and Lindhard, 1913). The majority of investigations have typically defined central command as a feedforward mechanism involving parallel activation of motor and cardiovascular centers (Goodwin et al., 1972). It has been proposed that central command may involve the simultaneous activation of two distinct networks, one for motor control and one for cardiovascular control (Williamson et al., 2006). While regions of the higher brain involved in the motor command component have been more readily and consistently identified (i.e. supplementary motor areas, premotor areas, and cerebellum), areas related to the cardiovascular component have proven more difficult to define. Of note, valuable studies exploring simulated central command using both animal and humans have more clearly identified the subcortical neural circuitry (Eldridge et al., 1985; Thornton et al., 2002; Green et al., 2007). The neural connectivity between these subcortical regions and higher brain centers has provided valuable insight regarding potential cortical site(s) from which the central command signal may originate. UT Southwestern Medical Center, Department of Health Care Sciences, 5323 Harry Hines Blvd., Dallas, TX , United States. Tel.: ; fax: address: jon.williamson@utsouthwestern.edu. It is well accepted that a central cardiovascular command works in conjunction with a peripheral feedback (e.g. metaboreflex) and baroreflex mechanisms to modulate the cardiovascular response to exercise. While this complex control system produces an effective and efficient cardiovascular response to exercise, it also makes the study of human central command more complicated. Many studies have examined cortical activation during exercise in humans. However, to identify the cortical origin of the central command signal one would need to design a study addressing several key issues. The magnitude of the exerciseinduced central command response would need to be quantified. This has most often been assessed using an individual's perception of effort, effort sense and/or heart rate change during exercise, independent of the actual workload or force production (Mitchell, 1990). The study would need to employ a brain mapping technology with sufficient spatial and temporal resolution to define specific cortical sites involved. Examples of this would include positron emission tomography (PET), single photon emission computed tomography (SPECT) and/or functional magnetic resonance imaging (fmri). Additionally, the central command signal would need to be activated independent of (or somehow uncoupled from) changes in muscle afferent feedback and baroreflex input. Given the challenges of meeting these criteria, only a handful of studies have attempted to identify the neuroanatomical correlates of a central command, but the findings have not yielded a definitive conclusion. These central command studies used a variety of methodologies to uncouple feedforward command signals from peripheral feedback, as well as brain mapping technologies, with the similar intent of identifying higher cortical structures involved in a central command. Nowak et al. (1999) employed regional anesthesia to block afferent feedback from the working limb during handgrip exercise. During exercise with anesthesia, they found an activation of the sensory motor area, supplementary motor area (SMA) and cerebellum, suggesting a feedforward network of motor command functioning independent of a sensory / 2014 Elsevier B.V. All rights reserved.

57 4 J.W. Williamson / Autonomic Neuroscience: Basic and Clinical 188 (2015) 3 4 feedback. Thornton et al. (2001) used hypnosis to contrast conditions of imagined downhill and uphill cycling, as well as non-exercise volitional hyperventilation condition. They also reported a motor-command activation (i.e. respiratory response) that involved the SMA, premotor area, and cerebellum. While providing evidence of anatomical regions capable of serving as motor command, these sites are not likely involved in cardiovascular regulation. Employing a slightly different approach to help better define areas of cortical cardiovascular control, Williamson et al. (2001) used hypnosis to change a perception of effort during a constant load cycling exercise. Contrasting the baseline exercise with the condition of perceived heavier workload (i.e. effort sense) they did not find differences in areas of motor control (based on the experimental design), but did report increased activation in the anterior cingulate cortex, insular cortex and thalamic regions; decreases in perceived effort decreased activity in these same regions. Williamson et al. (2002) also utilized hypnosis to compare patterns of cortical activation between actual and imagined handgrip exercise. They reported activation of the insular and anterior cingulate cortex and these regions appeared to be involved in cardiovascular modulation independent of any muscle afferent feedback. Nowak et al. (2005) investigated an attempted foot lifting in paraplegic subjects and reported activation in the cerebellum and insular cortex, with the insula being activated under conditions involving isolated feedforward control. Taken together, these studies found an activity of higher brain regions associated with motor control and cardiovascular regulation, but they did not necessarily account for the possibility of simultaneous changes in metaboreflex or baroreflex input. To examine this issue of metaboreflex input to cortical regions, Williamson et al. (2003) used post-exercise cuff occlusion (PECO) to compare patterns of brain activation between handgrip exercise and PECO. They reported insular activation during both exercise and PECO. However, there were differences in specific regions of insular activity, suggesting distinct regions within the insular cortex for central command and peripheral feedback. Sander et al. (2010) used a similar approach, but did not detect differences in patterns of insular activation between handgrip and PECO. They suggested that insular activity was related to blood pressure increases and/or ischemia-induced pain, as the insular is well-known to respond to both baroreflex and nociceptive stimuli. It has also been reported that a decreased perception of effort during exercise resulted in decreases in insular activity without concomitant changes in blood pressure (Williamson et al., 2002) suggesting that these changes were independent of baroreflex input. These differences might be attributed to several factors, one being that the insular cortex may serve multiple functions. As identified by Mitchell (1990), the relative importance of a central command and peripheral reflex components in determining cardiovascular responses to exercise is dependent upon the type of exercise (static or dynamic), the intensity of exercise, the time after onset of exercise (immediate, steady state, exhaustion, etc.), and the effectiveness of blood flow in meeting the metabolic needs of the contracting muscles. At present, there is no consensus for a cortical site (or higher brain network) of a central command serving as the origin of the signal that produces parallel activation of motor and cardiovascular centers. Moreover, the findings suggest that networks involving a central motor command and a central cardiovascular control can function independent of each other, further highlighting the complexities of a central command. The redundancy of exercise regulatory systems, feedback mechanisms and high degree of neural connectivity throughout the brain have proven challenging for investigators. As such, we are still left without a definitive neuroanatomy for a central command. Although new pieces of the central command puzzle have been provided, without further neurophysiological investigation, the total picture surrounding a central command and human exercise will remain incomplete. References Eldridge, F.L., Millhorn, D.E., Kiley, J.P., Waldrop, T.G., Stimulation by central command of locomotive and respiration and circulation during exercise. Respir. Physiol. 59, Goodwin, G.M., McCloskey, D.I., Mitchell, J.H., Cardiovascular and respiratory responses to changes in central command during isometric exercise at constant muscle tension. J. Physiol. 226, Green, A.L., Wang, S., Purvis, S., Owen, S.L.F., Bain, P.G., Stein, J.F., Guz, A., Aziz, T.Z., Paterson, D.J., Identifying cardiorespiratory neurocircuitry involved in central command during exercise in humans. J. Physiol. 578 (2), Krogh, A., Lindhard, J., The regulation of respiration and circulation during the initial stages of muscular work. J. Physiol. Lond. 47, Mitchell, J.H., Neural control of the circulation during exercise. Med. Sci. Sports Exerc. 22 (2), Nowak, M.K.S. Olsen, Law, I., Holm, S., Paulson, O.B., Secher, N.H., Command-related distribution of regional cerebral blood flow during attempted handgrip. J. Appl. Physiol. 86 (3), Nowak, M.S., Holm, F., Biering-Sorensen, N.H. Secher, Friberg, L., Central command and insular activation during attempted foot lifting in paraplegic humans. Hum. Brain Mapp. 25, Sander, M., Macefield, V.G., Henderson, L.A., Cortical and brain stem changes in neural activity during static handgrip and postexercise ischemia in humans. J. Appl. Physiol. 108, Shoemaker, J.K., Wong, S.W., Cechetto, D.F., Cortical circuitry associated with reflex cardiovascular control in humans: does the cortical autonomic network speak or listen during cardiovascular arousal. Anat. Rec. 295, Thornton, J.M., Guz, A., Murphy, K., Griffith, A.R., Pedersen, D.L., Kardos, A., Leff, A., Adams, L., Casadei, B., Paterson, D.J., Identification of higher brain centres that may encode the cardiorespiratory response to exercise in humans. J. Physiol. 533 (3), Thornton, J.M., Aziz, T., Schlugman, D., Paterson, D.J., Electrical stimulation of the midbrain increases heart rate and arterial blood pressure in humans. J. Physiol. 539 (2), Williamson, J.W., McColl, R., Mathews, D., Raven, P.B., Mitchell, J.H., Morgan, W.P., Hypnotic manipulation of effort sense during exercise: cardiovascular responses and brain activation. J. Appl. Physiol. 90, Williamson, J.W., McColl, R., Mathews, D., Raven, P.B., Mitchell, J.H., Morgan, W.P., Brain activation by central command during actual and imagined handgrip. J. Appl. Physiol. 92, Williamson, J.W., McColl, R., Mathews, D., Evidence for central command activation of the human insular cortex. J. Appl. Physiol. 94, Williamson, J.W., Fadel, P.J., Mitchell, J.H., New insights into central cardiovascular control during exercise in humans: a central command update. Exp. Physiol. 91,

58 Acta Physiol 2010, 199, REVIEW Cardiovascular function in the heat-stressed human C. G. Crandall 1,2 and J. González-Alonso 3 1 Department of Internal Medicine Institute for Exercise and Environmental Medicine, Texas Health Presbyterian Hospital Dallas, Dallas, TX, USA 2 University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA 3 Centre for Sports Medicine and Human Performance, Brunel University, West London, Middlesex, UK Received 14 July 2009, accepted 10 September 2009 Correspondence: C. G. Crandall, PhD, Institute for Exercise and Environmental Medicine, Presbyterian Hospital of Dallas, 7232 Greenville Ave., Dallas, TX 75231, USA. craigcrandall@texashealth. org; or J. González-Alonso, PhD, Centre for Sports Medicine and Human Performance, Brunel University, Uxbridge, Middlesex UB8 3PH, UK. j.gonzalez-alonso@brunel. ac.uk Both authors contributed equally to this work Abstract Heat stress, whether passive (i.e. exposure to elevated environmental temperatures) or via exercise, results in pronounced cardiovascular adjustments that are necessary for adequate temperature regulation as well as perfusion of the exercising muscle, heart and brain. The available data suggest that generally during passive heat stress baroreflex control of heart rate and sympathetic nerve activity are unchanged, while baroreflex control of systemic vascular resistance may be impaired perhaps due to attenuated vasoconstrictor responsiveness of the cutaneous circulation. Heat stress improves left ventricular systolic function, evidenced by increased cardiac contractility, thereby maintaining stroke volume despite large reductions in ventricular filling pressures. Heat stress-induced reductions in cerebral perfusion likely contribute to the recognized effect of this thermal condition in reducing orthostatic tolerance, although the mechanism(s) by which this occurs is not completely understood. The combination of intense whole-body exercise and environmental heat stress or dehydration-induced hyperthermia results in significant cardiovascular strain prior to exhaustion, which is characterized by reductions in cardiac output, stroke volume, arterial pressure and blood flow to the brain, skin and exercising muscle. These alterations in cardiovascular function and regulation late in heat stress/dehydration exercise might involve the interplay of both local and central reflexes, the contribution of which is presently unresolved. Keywords baroreflexes, cerebral perfusion, dehydration, exercise, hyperthermia. Humans have the capability to withstand large variations in environmental temperatures, while relatively small increases in internal temperature (i.e. as little of 3 C) can lead to injury and even death. Elevations in skin blood flow and sweating are the primary heat exchange mechanisms in humans that protect against a heat-related injury. The importance of these mechanisms is exemplified in the calculation that if heat was not liberated from skin, internal temperature would reach the upper safe limit within 10 min of moderate exercise (Kenney & Johnson 1992). These heat-dissipating responses are accompanied by important, even critical, cardiovascular adjustments, which, if they were not present, would compromise thermal regulation during exercise and/or exposure to elevated environmental temperatures. There is little doubt that Johannes Lindhard understood these important concepts. In fact, in 1910 he published an article titled Investigations into the conditions governing the temperature of the body in which the relationship between rectal temperature and heat-producing processes of the body was investigated (Lindhard 1910). Later he reported on the temperature of human muscles at rest and during exercise (Buchthal et al. 1944). A student who studied under Lindhard s Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 407

59 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, and Krogh s guidance, Marius Nielsen, continued to investigate human temperature regulation (Nielsen 1938, Christensen et al. 1942, Asmussen & Nielsen 1947, Nielsen & Nielsen 1962, 1965a,b). The objective of this article is to present findings pertaining to cardiovascular responses associated with human temperature regulation. The review is divided into two parts with part one focusing on cardiovascular responses to passive (i.e. non-exercising) heat stress, while the second part focuses on cardiovascular responses associated with exercise heat stress. Given numerous outstanding review articles on these topics (Rowell 1974, 1983, 1986a,b,c, Johnson 1992, Johnson & Proppe 1996, González-Alonso et al. 2008), this review will focus primarily on relatively new findings (i.e. within the prior 15 years), while salient work done prior to this period will be only briefly addressed. Key cardiovascular responses to passive heat stress Clearly the primary stimulus by which internal temperature is elevated in humans is through exercise. However, heat-related injuries and deaths can occur in nonexercising humans, especially those with prior medical conditions such as cardiovascular disease (Semenza et al. 1996, 1999). Particularly prevalent was the large number of excess deaths during the 1995 Chicago heat wave of individuals with a prior heart condition (Semenza et al. 1996). Investigations into cardiovascular responses to passive heat stress are valuable towards the identification of potential mechanisms responsible for heat-induced injuries and deaths in patient populations. Moreover, cardiovascular responses to passive heat stress provide a benchmark from which the effects of combined exercise and heat stress can be compared. Studies investigating thermal and cardiovascular responses to passive heat stress typically heat the subject via water-perfused suits, in which hot water is perfused through a tube-lined suit worn by each subject; by water immersion of the entire body or just limbs (such as the legs) in warm water; or by exposure of the individual to warm environmental conditions using a climatic chamber. In all cases an elevation in skin temperature is the primary mode by which internal temperature is elevated; although depending on the approach the magnitude of the increase in skin temperature can greatly differ. A limitation of the interpretation of these studies is the lack of consistency of the magnitude of the heat stress, primarily defined by the magnitude of the increase in internal temperature; with some studies reporting heat stress being as little as a 0.5 C increase in internal temperature whereas other studies report increasing internal temperature upwards to 2 C. During pronounced passive heat stress, skin blood flow of humans is estimated to increase from 300 ml min )1 upwards to 7500 ml min )1 (Rowell et al. 1969a, Rowell 1986c). In order to prevent large decreases in arterial blood pressure due to pronounced increases in total vascular conductance associated with cutaneous vasodilation, cardiac output must increase along with decreases in vascular conductance of noncutaneous beds. The combination of these responses results in either no change, or only minimal reductions in arterial blood pressure. Rowell et al. (1969a) and Rowell (1986c) showed that during a pronounced passive heat stress, cardiac output can be as high as 13 L min )1, with 50% of that value estimated to be directed towards skin. This elevation in cardiac output is primarily mediated through increases in heart rate as stroke volume either does not change or is marginally increased in young healthy heat-stressed subjects (Damato et al. 1968, Rowell et al. 1969a, Minson et al. 1998). Redistribution of blood flow from non-cutaneous tissues is evident from passive heat stress studies showing that post-ganglionic muscle sympathetic nerve activity increases (Niimi et al. 1997, Crandall et al. 1999a, 2003, Cui et al. 2002b,c, 2004a, Keller et al. 2006) while vascular conductance in other beds (i.e. splanchnic and renal) decreases (Rowell et al. 1971, Rowell 1986c, Minson et al. 1998). Thus, pronounced heat stress in humans has been termed a hyperadrenergic state as it is accompanied with the aforementioned responses (Rowell 1990). These primary cardiovascular responses to passive heat stress are depicted in Figure 1. Recent studies provide further insight into the effects of heat stress on cardiovascular responses, specifically the effects of heat stress on baroreflex responsiveness, central blood volume, cardiac function, and control of the cerebral vasculature. These areas are the primary focus of the first part of this review. Heat stress and baroreflex responsiveness The mechanisms responsible for reduced orthostatic tolerance in heat-stressed humans have not been fully elucidated, although Lind et al. (1968) showed this response was not due to enhanced pooling of blood in the leg during the orthostatic challenge. The central components governing thermoregulation are located in the hypothalamus (Strom 1960), and electrical stimulation of the hypothalamus modifies the baroreceptor reflex (Reis & Cuénod 1965, Gebber & Snyder 1970). Thus it seemed feasible that heat stress may impair baroreflex function, which could compromise blood pressure control while individuals are in this thermal condition. Based upon numerous studies evaluating the effects of heat stress on baroreflex responses, four key observa- 408 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

60 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system Figure 1 Classic cardiovascular responses to increases in skin temperature (T s ) resulting in a large increases in body temperature (T b ) reported as a per cent change in the indicated value relative to pre-heat stress baseline. CO, cardiac output; HR, heart rate; SV, stroke volume; CBV, central blood volume; AoMP, aortic mean arterial blood pressure; RAMP, right atrial mean blood pressure; TPR, total peripheral resistance. Note that the effects of heat stress on central blood volume have recently been shown to decrease as opposed to slight increases observed by these investigators (see Fig. 2). Figure from Rowell et al. (1969a); republished with permission from The American Physiological Society. tions can be summarized. (1) Generally, heat stress does not alter baroreflex control of heart rate (Crandall 2000, Yamazaki & Sone 2000, Wilson et al. 2001, Yamazaki et al. 2001, Cui et al. 2002b). The exceptions are from a few studies in which the change in heart rate during relatively small spontaneous oscillations in blood pressure is attenuated during heat stress (Crandall et al. 2000, Lee et al. 2003). It is possible that these latter observations are due to reduced cardiac vagal activity associated with heating (Crandall et al. 2000) given that when greater changes in baroreceptor loading were caused either mechanically (Crandall 2000, Yamazaki & Sone 2000) or pharmacologically (Wilson et al. 2001, Cui et al. 2002b), the baroreflex gain of the blood pressure to heart rate relationship was unchanged due to whole-body heating. (2) Baroreflex control of muscle sympathetic nerve activity is typically not attenuated by heat stress, but rather is unchanged or even elevated (Cui et al. 2002b, 2004a, Keller et al. 2006). The exception is a study showing attenuated changes in muscle sympathetic nerve activity relative to blood pressure changes induced by the valsalva manoeuvre in heat-stressed subjects (Yamazaki et al. 2003). Apart from the acute nature of the blood pressure change during the valsalva manoeuvre (i.e. typically less than 10 cardiac cycles), relative to the longer duration of the hypotensive challenge to systemic nitric oxide administration and lower body negative pressure, reasoning for these differing responses is not forthcoming. (3) Cutaneous vasoconstrictor responses are attenuated by local and whole-body heat stress (Wilson et al. 2002a), whereas muscle vasoconstrictor responses are not impaired when muscle temperature is elevated 4 C (Keller et al. 2007). Attenuated cutaneous adrenergic vasoconstrictor responsiveness (Wilson et al. 2002a), coupled with upwards to 50% of cardiac output being directed towards the skin in the heatstressed subject, may be the mechanism responsible for attenuated blood pressure changes upon changing carotid sinus transmural pressure (Crandall 2000), as well as during systemic phenylephrine administration (Cui et al. 2002c). (4) Despite somewhat varied baroreflex responses to heat stress, this exposure consistently shifts the baroreflex curve to the prevailing heart rate, muscle sympathetic nerve activity and blood pressure (Crandall 2000, Yamazaki et al. 2001, 2003, Cui et al. 2002b, Crandall et al. 2003, Keller et al. 2006). In summary, the bulk of the data do not support the hypothesis that heat stress attenuates baroreceptor control of heart rate or muscle sympathetic nerve activity to relatively pronounced changes in arterial blood pressure. However, heat-induced alterations in post-synaptic cutaneous vasoconstrictor responsiveness may attenuate baroreflex control modulation of systemic vascular conductance. Heat stress and blood volume distribution Passive heat stress causes pronounced reductions in central venous pressure which, depending on the magnitude of the heat stress, can approach 0 mmhg (Rowell et al. 1969a, Johnson & Proppe 1996, Minson et al. 1998, Crandall et al. 1999b, Peters et al. 2000, Wilson et al. 2007, Keller et al. 2009). The mechanism responsible for this reduction in central venous pressure Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 409

61 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, has been proposed to be due to a redistribution of blood from the central circulation to cutaneous vascular beds (Johnson & Proppe 1996). Despite heat stressinduced reductions in central blood volume being hypothesized (Müller 1905, Glaser et al. 1950, Eisalo 1956, Koroxenidis et al. 1961, Frayser et al. 1966), Rowell et al. (1969a) reported a slight increase in central blood volume during whole-body heat stress when indexed from mean transit time of a dye injected into the right atrium and sampled at the aortic arch (see Fig. 1). Such a finding is perplexing given the discord between reductions in central venous pressure and Rowell s observation of a slight increase in central blood volume. To clarify this question, Crandall et al. (2008) evaluated changes in regional blood volume during passive heat stress using technetium-99m labelled autologous red blood cells, coupled with gamma camera imaging. Although with this approach absolute measures of central blood volume were not obtained, relative changes in regional blood volumes were quantified. The heat stress caused typical haemodynamic responses, similar to that outlined in Figure 1, including a reduction in central venous pressure from to mmhg (P < 0.001). Accompanying this central venous pressure response were greater reductions in thoracic blood volume (14 2%), heart blood volume (18 2%) and the volumes of the heart plus the central vascular structures (17 2%), relative to responses in subjects who were not heat stressed but were supine for a similar duration following technetium-99m administration as the heat-stressed subjects (Fig. 2). In addition, heat stress decreased liver (23 2%) and spleen (27 2%) blood volume, which is consistent with previously reported reductions in splanchnic blood flow by heat stress (Rowell et al. 1969a, 1971, Minson et al. 1998). These data provide clear evidence that heat stress decreases central blood volume, which is likely due to a redistribution of blood from the central to the cutaneous vascular beds, coupled with increases in cardiac output (Johnson & Proppe 1996). Heat stress and cardiac function Consistent with a reduction in central venous pressure, passive heat stress results in parallel decreases in left ventricular filling pressure as indexed from pulmonary capillary wedge pressure (Wilson et al. 2007, 2009). Reduced central blood volume and ventricular filling pressures, accompanied with preserved or slightly elevated stroke volumes has led to the suggestion that heat stress increases the inotropic state of the heart (Rowell 1986c, Johnson & Proppe 1996). Consistent with that hypothesis, Crandall et al. (2008) showed that whole-body heat stress increases ejection fraction. (a) Change in blood volume (percent) (b) Change in blood volume (percent) P = Heat stress Time control Heat stress Time control P < P < Thorax Heart Heart & central vasculature P < P < P < Figure 2 Percent change in blood volume from the indicated regions between experimental (i.e. heat stressed) and time control subjects. In each of the indicated regions heat stress significantly reduced blood volume relative to the time control trials. Figure from Crandall et al. (2008); republished with permission from Wiley-Blackwell. However, these findings must be interpreted with the recognition of the indirect nature of this measurement and the load dependency of ejection fraction, resulting in an imprecise measure of systolic function, particularly given changes in ventricular loading status during a heat stress. For this reason, Brothers et al. (2009a) identified the effects of heat stress on echocardiographic indices of cardiac systolic and diastolic function. The primary observations from that study were: (1) Heat stress has no effect on indices of diastolic function as indicated by an unchanged left ventricular filling velocity, an unchanged early diastolic mitral annular velocity and an unchanged ratio of blood velocity/mitral annular velocity during the early phase of diastole. However, the preservation of these indices of diastolic function, despite heat stress-induced decreases in ventricular filling pressures (Rowell et al. 1969a, Minson et al. 1998, Crandall et al. 1999b, Wilson et al. 2007, 2009) and central blood volume (Crandall et al. 2008), leaves room for speculation that diastolic function is perhaps improved during heat stress. (2) Heat stress increased left ventricular systolic 410 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

62 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system function as evidenced by an increase in peak septal and lateral mitral annular systolic velocities and isovolumic acceleration (Fig. 3). (3) Heat stress significantly increased left atrial systolic function as evidenced by an increased velocity of blood during the atrial contraction phase of left ventricular diastolic filling and an increased velocity of the septal and lateral mitral annulus during the late phase of diastolic relaxation relative to normothermia. Increases in left ventricular end-diastolic cross-sectional diameter, often indicated by its filling pressure, increase the ability of the left ventricle to produce force and thereby stroke volume (Sarnoff & Mitchell 1961, Katz 2006). This concept, termed Frank Starling mechanism (Frank 1895, Patterson & Starling 1914, Patterson et al. 1914), is represented in vivo by a series of hyperbolic curves relating changes in pulmonary capillary wedge pressure (as an index of left ventricular end-diastolic volume) relative to stroke volume. Given the hyperbolic shape of these curves, the magnitude of change in stroke volume for a given change in filling pressure is largely affected by the shape of the curve and the location of the operating point on that curve. A reduction in left ventricular filling pressure caused by passive heat stress, coupled with an absence of a reduction in stroke volume, suggests heat stress causes a leftward shift of the Frank Starling curve. Wilson et al. (2009) experimentally validated this hypothesis upon examining the relationship between pulmonary capillary wedge pressure and stroke volume during 15 and 30 mmhg lower body negative pressure while subjects were normothermic and heat stressed (Fig. 4). Heat stress shifted the operating point to a steeper portion of the Frank Starling curve such that for a given reduction in ventricular filling pressure there was a greater reduction in stroke volume. This larger reduction in stroke volume was not compensated for by a proportionally greater increase in heart rate, resulting in a larger reduction in cardiac output for a given reduction in left ventricular filling pressure or level of lower body negative pressure when the individual was heat stressed (Wilson et al. 2009), which likely is a primary mechanism for heat stress-induced reductions in orthostatic tolerance (Keller et al. 2009). (a) 20 Septal mitral annular S 20 Tissue velocity (cm s 1 ) P = P = Tissue velocity (cm s 1 ) 0 NT WBH NT WBH NT WBH NT WBH 0 (b) 5 Septal mitral annulus Lateral mitral annulus 5 Isovolumic acceleration (m s 2 ) P = 0.03 P < Isovolumic acceleration (m s 2 ) 0 NT WBH NT WBH NT WBH NT WBH 0 Figure 3 Peak septal and lateral mitral annular systolic velocities (S ; panel a) and isovolumic acceleration of the septal and lateral mitral annulus (panel b). Individual (left-hand side of each panel) and group averaged (right-hand side of each panel) echocardiographic measurements of the indicated data during normothermic (NT) and whole-body heat-stress (WBH) conditions. Increases in the indicated parameters by heat stress are indicative of an increase in cardiac systolic function. Figure from Brothers et al. (2009a); used with permission from The American Physiological Society. Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 411

63 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, Stroke volume (ml) Normothermia Heat stress Pulmonary capillary wedge pressure (mmhg) Figure 4 Effect of thermal stress on the Frank Starling relation via plotting the relation between pulmonary capillary wedge pressure and stroke volume. Data points were generated via lower body negative pressures (LBNP) of 0, 15 and 30 mmhg. The arrows indicate the operating point for the respective thermal conditions. The operating point is defined as the prevailing pulmonary capillary wedge pressure and stroke volume prior to the onset of LBNP. Figure from Wilson, Brothers, Tollund, Dawson, Nissen, Yoshiga, Jons, Secher and Crandall. J Physiol 587, , 2009; republished with permission from Wiley-Blackwell. Heat stress and the cerebral vasculature Whole-body heat stress decreases cerebral perfusion in the supine resting human (Wilson et al. 2002b, 2006, Fan et al. 2008, Fujii et al. 2008, Low et al. 2008, 2009, Brothers et al. 2009b), which is not entirely accounted for by concurrent reductions in arterial carbon dioxide tension (Wilson et al. 2006, Brothers et al. 2009b). The latter point was recently confirmed given that cerebral perfusion and cerebral vascular conductance remained well below pre-heat stress levels after end-tidal carbon dioxide partial pressures, which can be reduced by 15+ mmhg during heating (Wilson et al. 2006, Crandall et al. 2008, Fan et al. 2008, Fujii et al. 2008, Brothers et al. 2009b), were returned to pre-heat stress levels (Brothers et al. 2009b) (Fig. 5). The mechanism by which heat stress decreases cerebral vascular conductance and perfusion independent of reduced arterial carbon dioxide partial pressures is not clear. One possibility could be increased cerebral sympathetic activity. Sympathetic nerve activity increases by approximately 90% to the muscle vasculature and % to the skin vasculature during passive heat stress (Bini et al. 1980, Niimi et al. 1997, Cui et al. 2002a, 2004b,c, Keller et al. 2006). Furthermore, blood flow to the renal and splanchnic regions is reduced presumably as a result of increased sympathetically mediated vasoconstriction (Rowell et al. 1970, 1971, Johnson & Proppe 1996, Minson et al. 1998). While the role of sympathetic control of the cerebral vasculature remains controversial (van Lieshout & Secher 2008, Strandgaard & Sigurdsson 2008), animal studies have shown that the cerebral arteries are richly innervated with sympathetic nerve fibres (Nelson & Rennels 1970, Edvinsson 1975). Evidence for cerebral sympathetic activity has also been demonstrated in human studies that have identified a reduction in cerebral perfusion during unilateral trigeminal ganglion stimulation (Visocchi et al. 1996) and an increase in cerebral perfusion after stellate ganglionic blockade (Umeyama et al. 1995, Ide et al. 2000). Furthermore, cerebral autoregulation is impaired following removal of autonomic neural activity with trimethaphan (Zhang et al. 2002) and systemic blockade of a 1 -adrenergic receptors with prazosin (Ogoh et al. 2008). Therefore, it is plausible to speculate that heat stress-induced increases in cerebral sympathetic activity may contribute to reductions in cerebral perfusion in the heatstressed human. From the aforementioned findings, it is clear that even prior to the onset of an orthostatic event cerebral perfusion is compromised by whole-body heat stress. Reduced cerebral perfusion in the heat-stressed human will attenuate the reserve by which cerebral blood flow can further decrease prior to the onset of syncope. Thus, heat-stress induced decreases in cerebral perfusion likely contribute to reductions in orthostatic tolerance. Maintenance of cerebral perfusion over a wide range of systemic blood pressures (e.g. from 60 to 150 mmhg) is carried out by a variety of mechanisms including cerebral autoregulation which acts to offset changes in perfusion pressure by adjusting the resistance of the cerebral vasculature (Heistad & Kontos 1983, Paulson et al. 1990). Should heat stress attenuate cerebral autoregulation, this would result in a greater reduction in cerebral perfusion for a given reduction in perfusion pressure, ultimately leading to reduced orthostatic tolerance. Consistent with this hypothesis, Wilson et al. (2006) identified that for a given level of orthostatic stress (via lower body negative pressure) when subjects were heat stressed the reduction in cerebral vascular conductance and perfusion was significantly greater, relative to when subjects were normothermic. Doering et al. (1999) were the first to seek whether heat stress alters cerebral autoregulation. Counter to an expected reduction, in mildly heat-stressed subjects (0.4 C increase in internal temperature) they reported an increase in an index of cerebral autoregulation. Given the relatively minor heat stress in that study, Brothers et al. (2009c) and Low et al. (2009) further investigated this question, using varied techniques in more profoundly heat-stressed subjects relative to the level of heating in the Doering et al. study. Regardless of whether the blood pressure oscillations were relatively small (Low et al. 2009) or quite large (Brothers et al. 2009c), cerebral 412 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

64 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system (a) End tidal carbon dioxide tension (mm Hg) Normothermia * Heat stress Heat stress + PETCO 2 clamp (b) Normothermia * Heat stress * Heat stress + PETCO 2 clamp Cerebral blood velocity (cm s 1 ) Figure 5 End-tidal carbon dioxide tension and middle cerebral artery blood velocity (MCA V mean ) during normothermia, heat stress, and heat stress after end-tidal carbon dioxide (PETCO 2 ) concentration was returned to pre-heat stress levels. The reduction in PETCO 2 concentration during heat stress was completely abolished by the PETCO 2 clamping procedure (panel a). Heat stress reduced MCA V mean relative to normothermia. Restoration of PETCO 2 to the normothermic level while subjects were heat stressed (heat stress + clamp) attenuated the decrease in MCA V mean relative to control heat stress without the clamp; however MCA V mean remained reduced when compared with normothermia (panel b). These data indicate that mechanisms other than reduced PETCO 2 contribute to the reduced cerebral perfusion that occurs in heat-stressed individuals. *Significantly different relative to normothermia; significantly different relative to control heat stress. Figure from Brothers et al. (2009b); republished with permission from Wiley-Blackwell. autoregulation was not compromised by heat stress. In contrast, and depending on the frequency of blood pressure oscillation analysed, these investigators found either no change or enhanced cerebral autoregulation by heat stress; the latter being consistent with the original findings of Doering et al. (1999). Taken together, wholebody heat stress either does not change or perhaps improves cerebral vascular autoregulation. Cardiovascular responses to combined exercise and heat stress The combination of exercise and heat stress can pose one of the most severe challenges to the regulation of the cardiovascular system in humans. The interplay between the magnitude of thermal stress, the intensity and duration of exercise and the individual training, heat acclimatization and hydration status will dictate the extent of the cardiovascular challenge. These factors should therefore be thoroughly considered in predicting whether heat stress will promote compensatory adjustments or severe circulatory strain that could lead to accelerated fatigue or collapse in exercising people. Conceivably, the greatest cardiovascular strain and alterations in cardiovascular regulation are expected to occur in untrained, unacclimated and hypohydrated subjects who perform intense, prolonged exercise in a hot and humid environment. World-class, heat-acclimated and euhydrated athletes might be on the other end of a continuum and show minimal alterations in cardiovascular function under comparable extreme exercise and environmental conditions. This second component of the review will primarily discuss evidence and ideas concerning the responses and regulation of the human cardiovascular system when confronted with the combined stresses of exercise and environmental heat stress or dehydration-induced hyperthermia. The impact of alterations in limb blood flow on convective heat transfer and internal temperature regulation will also be briefly discussed in the context of the classic observations of Nielsen (1938). Over the past 45 years our understanding of human cardiovascular function during thermal stress and exercise has been influenced very profoundly by the work of Loring Rowell and colleagues at the University of Washington. In a landmark study, they (Rowell et al. 1966) were the first to demonstrate in a group of untrained individuals performing graded exercise in a hot and a thermoneutral environment that severe thermal stress readily perturbs cardiovascular function when the intensity of exercise is moderate to intense. The key features indicative of cardiovascular strain during exercise were the significantly lower stroke volume, central blood volume, aortic pressure and cardiac output and the markedly elevated heart rate reaching near-maximal values. The decreased central blood volume with heat stress across all four submaximal exercise intensities examined supported the idea that a lowering in venous return and filling of the heart might be an important factor involved in the depressed stroke volume and ultimately cardiac output with combined heat stress and exercise. In line with this view, they showed in a subsequent study using a water-perfused suit that the marked alterations in central blood volume, central Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 413

65 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, venous pressure, stroke volume, heart rate and cardiac output evoked by superimposing severe heat stress upon light and high intensity cycling were almost restored to normothermic control values by perfusing the suit with cold water (Rowell et al. 1969b). During walking in a very hot environment, however, cardiac output increased over time due solely to increases in heart rate (Rowell et al. 1967). Thus this pioneer research demonstrated that the repercussions of combined heat stress and exercise on cardiovascular function are dependent upon the intensity of exercise and implied that cardiac output and perfusion pressure are compromised in untrained subjects when both the metabolic and thermoregulatory demands are high. Limb muscle and skin perfusion during combined exercise and heat stress These early observations of a compromised cardiovascular function during heat stress exercise and the latter finding that forearm blood flow (index of skin blood flow in the resting forearm) increases progressively during leg exercise (Johnson & Rowell 1975) form the basis for the most influential hypothesis in this field postulating that the combination of heat stress and exercise results in a competition between the exercising skeletal muscles and the skin compartments for the available cardiac output such that blood flow to the active muscles would be reduced at the expense of an elevated skin circulation (Rowell 1974, 1983). According to this hypothesis, the pumping capacity of the heart cannot meet the joint demands for blood flow of the exercising muscle and the skin during exercise and heat stress. Experimental evidence in animals and humans, however, shows that blood flow to active limb muscles and tissues is either maintained or increased when heat stress is superimposed upon light to moderate intensity prolonged exercise (Laughlin & Armstrong 1983, Armstrong et al. 1987, Savard et al. 1988, McKirnan et al. 1989, Nielsen et al. 1990, 1993, 1997). In the human studies, the cardiovascular system appeared to respond adequately to the additional thermoregulatory demand for an elevated skin perfusion by increasing cardiac output (Savard et al. 1988, Nielsen et al. 1990, 1993, 1997) and possibly reducing visceral blood flow beyond the control exercise levels (Rowell et al. 1965, Ho et al. 1997). To understand this response, it is important to bear in mind that cutaneous vasodilatation is noticeably restrained during combined heat stress and exercise compared with the levels observed during resting hyperthermic conditions (Johnson 1992, Kenney & Johnson 1992). During upright exercise, skin perfusion in the resting forearm reaches a plateau at a core temperature of 38 C despite further increases in skin and internal temperature (Brengelmann et al. 1977, González-Alonso et al. 1999c). Although knowledge of the skin perfusion across all body segments including the exercising limbs is very limited, the magnitude of the cardiac output increase and the reduction in visceral blood flow during heat stress exercise indirectly suggest that whole-body skin blood flow might be elevated by 1 2 L min )1 above control exercise (Nielsen et al. 1993, 1997, Ho et al. 1997, González-Alonso et al. 2000a). This estimate is based on the still unresolved concept that hyperthermia-induced muscle vasodilatation does not contribute to heat stress-mediated hyperaemia. Regardless of this possibility, evidence from submaximal exercise studies does not seem to support the hypothesis of a regulatory priority of the skin circulation over the skeletal muscle circulation whereby flow from active muscles is redistributed to the skin. It is universally accepted that severe heat stress suppresses VO 2max and work capacity during exhaustive incremental exercise (Rowell 1974, Hales et al. 1996, Sawka & Coyle 1999). The question then arises as to whether blood flow to active muscles is indeed reduced at the expense of an elevated skin perfusion when heat stress is superimposed upon exercise requiring maximal cardiovascular function and aerobic capacity. Surprisingly, direct data are still lacking on the blood flow responses of the active limb muscles and skin and the systemic circulation during incremental whole-body exercise in heat-stressed individuals. Notwithstanding, an integrative view of the exercising limb and systemic circulatory responses to maximal exercise was recently reported in trained individuals performing constant load cycling under heat stress and normal conditions (González-Alonso & Calbet 2003, González-Alonso et al. 2004). An advantage of this exercise model is that it allows investigation of the functional and regulatory capacity of the cardiovascular system in conditions where workload-related changes in metabolic demand and active muscle mass are minimal in comparison with that occurring during exhaustive incremental exercise or self-selective pacing trials. The potentially confounding influence of altered workload-related metabolic and mechanical reflexes on cardiovascular control is therefore minimized. In using this model, González-Alonso & Calbet (2003) showed that blood flow to the exercising legs increases similarly with and without heat stress during the first minute of maximal cycling but thereafter is attenuated and drops faster with heat stress accompanying a quicker decline in cardiac output and arterial blood pressure (Fig. 6). Strikingly, vascular conductance in the active legs and systemic circulation do not decline despite the concomitant increases in circulating catecholamines indicative of enhanced sympathetic vasoconstrictor activity. Importantly, systemic blood flow and arterial pressure responses are elevated or unchanged with heat stress compared with control 414 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

66 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system Cardiac output and two-legged blood flow (L min 1 ) Mean arterial blood pressure (mmhg 1 ) Systemic and two-legged vascular conductance (ml min 1 mmhg 1 ) Heat stress-maximal Normal Normal-maximal Time (min) conditions during the early stages of constant maximal exercise (González-Alonso & Calbet 2003). This contrasts with the diminished cardiac output and arterial pressure seen during moderate and intense exercise in untrained individuals (Rowell et al. 1966). It therefore seems that during the early stages of maximal heat stress exercise in trained individuals the higher thermoregulatory demand for skin blood flow is met at least in part by a 1 2 L min )1 higher cardiac output. However, hyperthermia more quickly pushes the cardiovascular system * * Figure 6 Haemodynamics during maximal whole-body exercise in heat-stressed humans. Systemic and exercising limb blood flow and vascular conductance during constant maximal exercise with heat stress and control conditions. Note the significant reductions in cardiac output and leg blood flow and arterial blood pressure leading to unchanged vascular conductance. *Significantly lower than corresponding peak exercise values, P < Significantly lower than control (normal) trials, P < Figure from González-Alonso & Calbet (2003); republished with permission from the American Heart Association. * * * * to its regulatory limit, where cardiac output and blood flow to exercising limb muscles and skin cannot be maintained for a longer duration. The mechanisms underlying the restrictions in locomotor limb blood flow during maximal exercise have not been systematically investigated, but they plausibly involve the interplay of local and central reflexes signalling alterations in thermal, metabolic, mechanical, barosensitive and vascular events in different regions of the body including the skeletal muscle, brain and heart (Rowell 2004, Mortensen et al. 2008). The reductions in locomotor limb blood flow and cardiac output during constant maximal exercise and the levelling off in these variables during incremental maximal exercise are temporally related (González-Alonso & Calbet 2003, Mortensen et al. 2005, 2008, Calbet et al. 2007, Vogiatzis et al. 2009). During constant load maximal exercise with and without heat stress, arterial and central venous pressures decline such that the vascular conductance of the active legs and systemic circulation remain unchanged. This suggests that active limb blood flow is reduced secondary to the decline in perfusion pressure, rather than an actual vasoconstriction of the vasculature perfusing the active muscle and skin. In this construct, the reduction in cardiac output might be indirectly involved in this process via its effect on perfusion pressure. Locally, the unaltered leg vascular conductance takes place in the presence of an everincreasing plasma noradrenaline concentration suggesting that neurally mediated vasoconstriction does not occur in the vessels perfusing the exercising limb muscles and skin, even though muscle and skin sympathetic vasoconstrictor activity are likely to be augmented (Ray & Gracey 1997, Ichinose et al. 2008). An enhanced metabolic vasodilatation evoked by accumulation of vasodilator substances in the blood (including ATP and other adenine nucleotides) and the muscle interstitium might offset the increased sympathetic vasoconstrictor drive at least in skeletal muscle thereby maintaining limb muscle vascular conductance (Vanhoutte et al. 1981). This phenomenon resembles the functional sympatholysis occurring in the skeletal muscle vasculature in conditions of increased sympathetic nerve drive during exercise and hypoxia (Remensnyder et al. 1962, Hanada et al. 2003, Rosenmeier et al. 2004), which underlies the maintenance of resting limb perfusion and the increase in exercising limb blood flow during hypoxic exercise. Cardiac function during combined exercise and heat stress Our knowledge and understanding of cardiac function during intense exercise and of the mechanisms underpinning the fall in stroke volume prior to exhaustion are still inadequate. From a cardiac perspective, the fall in Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 415

67 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, stroke volume and cardiac output during constant maximal exercise with and without exogenous heat stress might be the result of a number of factors that negatively affect cardiac pre-load, ventricular afterload and/or myocardial contractility (Rowell 1974, Poliner et al. 1980, Higginbotham et al. 1986). Neither blood pooling in the compliant cutaneous vasculature (which could potentially diminish venous return) nor augmented ventricular afterload appear likely possibilities as stroke volume, central venous pressure and arterial pressure fall similarly during heat stress and control conditions (González-Alonso et al. 2000a, González- Alonso & Calbet 2003). The decline in central venous pressure could be interpreted to mean that a reduction in venous return contributes to the fall in stroke volume (Rowell et al. 1966, Rowell 1974). However, a reduced venous return and a concomitant diminution in left ventricular pre-load do not seem to exert an independent effect because stroke volume only declines during the last minute of exercise, but right atrial pressure declines from the start of exercise. This suggests that several factors interact to transiently depress diastolic and/or systolic cardiac function during maximal exercise. Seeing the declines in cardiac output, exercising limb blood flow and brain circulation, it seems possible that the coronary circulation and left ventricular function are transiently suppressed during maximal exercise, thereby contributing to the reduction in stroke volume. A small attenuation in the myocardial perfusion-to-work relationship could lead to myocardial dysfunction, as oxygen (extraction) reserve is very small to compensate for a significant blunting in oxygen supply. Severe tachycardia might be another factor. Studies in humans and dogs manipulating heart rate by pacing the heart demonstrate that severe tachycardia leads to disproportional reductions in diastolic filling time and left ventricular enddiastolic volume which compromise stroke volume and cardiac output (Templeton et al. 1972, Parker & Case 1979). Human studies demonstrate that hyperthermiainduced tachycardia reduces stroke volume during exercise (González-Alonso et al. 1997, Fritzsche et al. 1999) and that blunting the increase in internal temperature and heart rate restores most of the fall in VO 2max evoked by marked hyperthermia and dehydration (Nybo et al. 2001). Taken together, the decline in stroke volume during maximal exercise with and without exogenous heat stress is associated with reduced venous return, severe tachycardia and possibly a blunting in myocardial oxygen supply in relation to actual cardiac work. Cerebral perfusion during combined exercise and heat stress The active limb muscles are not the only tissues that might experience a reduction in perfusion during exercise and heat stress. The human brain circulation might also be compromised (Nybo & Nielsen 2001, Nybo et al. 2002, González-Alonso et al. 2004). In this regard, middle cerebral blood velocity declines significantly during prolonged exercise in the heat while it is kept constant during exercise in a thermoneutral environment (Nybo & Nielsen 2001, Nybo et al. 2002). During constant maximal exercise, however, cerebral perfusion declines after 90 s, regardless of the presence or absence of heat stress, as indicated by a progressive drop in both middle cerebral artery mean blood velocity and frontal cortex tissue oxygenation and the concomitant increases in brain oxygen extraction (González-Alonso et al. 2004) (Fig. 7). These responses are in sharp contrast to that happening during the first 90 s of exercise where middle cerebral artery mean blood velocity increases and cerebral oxygen extraction and frontal cortex tissue oxygenation remain stable. This suggests that global brain aerobic metabolism increases early in exercise possibly in response to enhanced neural activation in regions of the brain related to locomotion, the maintenance of equilibrium, vision and cardiovascular control (Delp et al. 2001). The large increases in oxygen extraction that occur after 90 s of maximal exercise are accompanied by smaller reductions in cerebral perfusion, signifying that global brain metabolism and neural drive is further enhanced when approaching exhaustion. Therefore, the physiological repercussions of reductions in perfusion to the brain are apparently less severe than in contracting skeletal and cardiac muscle because, in contrast to the muscles exhausted oxygen reserve, the human brain maintains a large oxygen reverse on exhaustion, which appears to protect this vital organ against the small declines in oxygen delivery occurring during exercise. The suppression of brain blood flow early in upright exercise is related to factors other than cardiac output because middle cerebral blood velocity declines when cardiac output is increasing during exercise (Nybo et al. 2002, González-Alonso et al. 2004). The blunted perfusion pressure is a more likely factor because the decreases in left and right middle cerebral artery blood velocity are temporally associated with reductions in arterial and central venous pressures (González-Alonso et al. 2004). A role of perfusion pressure on brain circulation is demonstrated during an orthostatic challenge where middle cerebral artery blood velocity declines drastically when arterial and central venous pressures are compromised (Van Lieshout et al. 2003). The decline in P a CO 2 associated with hyperthermiainduced hyperventilation may also be a factor accounting for a portion of the decline in cerebral blood flow during hyperthermic exercise (Rasmussen et al. 2006). Another all-encompassing possibility is that local 416 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

68 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system both vasodilator and vasoconstrictor activities are elevated. Clearly the elucidation of the local mechanisms involved in the decline in cerebral perfusion during prolonged and maximal heat stress exercise warrants further studies quantifying the contribution of the vasodilator and vasoconstrictor systems. Cardiovascular strain during exercise with dehydration and hyperthermia Figure 7 Cerebral circulation and oxygenation during maximal whole-body exercise in heat-stressed humans. Left and right middle cerebral artery blood velocity and near-infrared spectroscopy-determined cerebral tissue oxygenation at rest, during submaximal and maximal cycling and during 10 min of recovery in heat stress and control conditions. Note the marked reductions in blood velocity accompanying the declines in tissue oxygenation. *Higher than value at start of exercise, P < Lower than peak value during maximal exercise, P < From González-Alonso et al. (2004); republished with permission from Wiley-Blackwell. factors reducing the vasodilator and/or increasing the vasoconstrictor activities suppress brain perfusion. In this regard, the plasma concentration of the potent vasodilator adenosine triphosphate is elevated in the jugular vein (accompanying decreases in venous oxygen saturation), while the brain is apparently taking up large amounts of catecholamines on exhaustion in both conditions and the arterial and jugular venous carbon dioxide partial pressure is declining, suggesting that Cardiovascular responses to exercise in the heat with and without dehydration have been extensively studied (Saltin & Stenberg 1964, Sawka et al. 1979, Nadel et al. 1980, Montain & Coyle 1992a,b, González- Alonso et al. 1995, 1997, 1998, 1999c, 2000a). Different experimental approaches that reduce body water either prior to exercise (e.g. diuretics, sauna, exercise, water restriction) or during exercise (fluid restriction) have been used to investigate the effects of reduced body fluids on cardiovascular function. The cardiovascular strain produced by different methods of body water deficits is essentially similar (Sawka & Coyle 1999). However, factors such as environmental conditions, intensity, position and type of exercise all influence the extent of the dehydration-mediated cardiovascular alterations (Montain et al. 1998, González-Alonso et al. 1999c, 2000a). The cardiovascular strain inflicted by environmental heat stress and that evoked by dehydration bear several similarities probably because internal body hyperthermia and reduced body fluids are common elements accompanying both stresses. A major difference, however, is that skin temperature in the exercise-induced dehydration studies (usually performed in compensable hot environments with fan cooling) is normally maintained at C (Montain & Coyle 1992b, González-Alonso et al. 1999a), while in the heat stress studies using a water-perfused suit or no fan cooling it might gradually increase up to C (Rowell et al. 1969b, González-Alonso et al. 1999c). The progressive dehydration incurred during prolonged moderate intensity cycling in compensable heat conditions is associated with gradual reductions in perfusion pressure and blood flow to the skin and locomotor limb tissues accompanying increases in plasma noradrenaline levels and core temperature (Sawka et al. 1979, Montain & Coyle 1992a,b, González-Alonso et al. 1995, 1998, 2000a) (Fig. 8). Cardiac output also declines with marked dehydration and hyperthermia because the larger declines in stroke volume compared with the parallel increases in heart rate (Sawka et al. 1979, Montain & Coyle 1992a,b, González-Alonso et al. 1997). Brain and visceral perfusion are likely to also decline as the lowering in exercising leg blood flow and skin blood flow account Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 417

69 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, (a) Cardiac output (L min 1 ) (b) Two-legged blood flow (L min 1 ) (c) Blood flow non-exercising tissues (L min 1 ) (d) Forearm blood flow (ml (100 g) 1 min 1 ) * Dehydration Control * * * * * Time (min) Figure 8 Haemodynamics with dehydration during prolonged exercise in the heat. Systemic and peripheral blood flow during prolonged cycling in the heat with and without dehydration and hyperthermia. Note that the declines in cardiac output are accompanied by reductions in blood flow to the exercising legs, the skin and possible visceral blood flow. *Significantly lower than 20 min value, P < Significantly lower than control, P < From González-Alonso et al. (1998); republished with permission from Wiley-Blackwell. for two-thirds of the decline in cardiac output (González-Alonso et al. 1998) and similar levels of hyperthermia and global cardiovascular instability have been associated with reductions in the middle cerebral artery blood velocity and renal and splanchnic blood flow (Rowell et al. 1965, Nybo & Nielsen 2001). Remarkably, these alterations in cardiovascular function are completely prevented when people fully maintain hydration status by fluid ingestion during exercise in the heat (González-Alonso et al. 1995, 1998) or when dehydrated subjects exercise in the cold and intravascular fluid losses are restored with a plasma volume expander (González-Alonso et al. 1997). The diminished intravascular fluids (or related haematological changes) and hyperthermia are therefore important factors underpinning the cardiovascular strain produced by dehydration during exercise in the heat. The reduction in stroke volume underlies the fall in cardiac output with dehydration and hyperthermia (Montain & Coyle 1992a, González-Alonso et al. 1997, 1999b, 2000a). Interestingly, when thermoregulatory demands of exercise are minimized in a cold environment and the rise in core temperature is blunted, the decline in stroke volume and the increase in heart rate are significantly attenuated. In these conditions the significant drop in cardiac output normally observed with dehydration and hyperthermia in the heat is prevented (González-Alonso et al. 2000a). Plasma volume expansion in dehydrated and normothermic subjects completely reverses the otherwise small reduction in stroke volume during exercise in the cold, despite the persistent 3 4 L extravascular dehydration (González- Alonso et al. 1997). Similarly, hyperthermia without dehydration induces a small decline in stroke volume during exercise in the heat, which is associated with an increased heart rate; whereas preventing dehydration and hyperthermia via fluid ingestion fully ameliorates these alterations in both stroke volume and heart rate (González-Alonso et al. 1995, 1997, 1998). Hence the reductions in cardiac stroke volume underlying the decline in cardiac output in dehydrated and hyperthermic individuals might be largely related to diminished intravascular volume and the hyperthermia-induced tachycardia. In 1938, Marius Nielsen (who worked closely with Lindhard and Krogh in The Zoophysiological Laboratory at the University of Copenhagen) demonstrated that core temperature during prolonged moderate exercise was similar across a wide range of ambient temperatures (i.e C) (Nielsen 1938). Because heat exchange pathways are very different in cold and hot environments, Nielsen s seminal observation provided the foundations for the idea that core temperature and thus the underlying heat fluxes between the exercising limb muscles and the environment surrounding the limbs, the exercising limbs and the body core, and the body core and the environment surrounding the trunk are well-regulated responses. In this context, the aforementioned dehydration-induced reductions in exercising limb blood flow might impair convective heat transport from active muscle to the surrounding environment and the body core and thus contribute to hyperthermia, independently of the influences of concurrent reductions in sweating rate and skin blood flow (Nadel et al. 1980, Montain & Coyle 1992a). 418 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

70 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system This idea was put to test by quantifying the convective heat transfer from the exercising leg to the body core (i.e. leg blood flow arterial venous (a v) temperature difference, according to the Fick principle) and total heat production from the heat equivalent of leg VO 2 (González-Alonso et al. 1999a, 2000b). The results suggested that dehydration impairs heat transfer from the leg to core. More strikingly, they indicated that more than onehalf of the metabolic heat liberated in the contracting leg muscles is dissipated directly to the environment surrounding the leg. These findings highlight the importance of investigating heat-dissipating mechanisms in the exercising limbs and how the stresses of dehydration and heat stress impact upon thermoregulatory function. Conclusions and future directions Passive heat stress has the capacity of causing pronounced strain on the cardiovascular system, evidenced by large increases in sympathetic neural activity, heart rate and left ventricular contractility, coupled with reductions in central blood volume, left ventricular filling pressures and cerebral perfusion. The mechanisms resulting in compromised blood pressure control that accompanies such exposure are not entirely clear, although the prevailing data do not support the hypothesis of heat stress-induced impairment in baroreflex responsiveness. Compelling evidence indicates that the combination of intense whole-body exercise and environmental heat stress or dehydration induced-hyperthermia results in significant cardiovascular strain prior to exhaustion, which is characterized by reductions in cardiac output, stroke volume, arterial pressure and blood flow to the brain, skin and exercising muscle. The local and central mechanisms underpinning these responses remain unresolved. Reductions in skeletal muscle blood flow might not only affect muscle metabolism during exercise but also convective heat transfer to the environment surrounding the exercising limbs and the body core. The study of the muscle and skin circulations in the exercising and non-exercising limbs and the heat fluxes within the body could provide novel insight into how core temperature and its effector responses are regulated in exercising humans. Conflict of interest There is no conflict of interest. This work was supported by the National Institutes of Health HL61388 and HL84072 (C.G.C.), Marie Curie Training program, Gatorade Sports Science Institute, Team Denmark, Novo Nordisk and the Danish National Research Foundation (J.G.-A.). References Armstrong, R.B., Delp, M.D., Goljan, E.F. & Laughlin, M.H Progressive elevations in muscle blood flow during prolonged exercise in swine. J Appl Physiol 63, Asmussen, E. & Nielsen, M The regulation of the bodytemperature during work performed with the arms and with the legs. Acta Physiol Scand 14, Bini, G., Hagbarth, K.E., Hynninen, P. & Wallin, B.G Thermoregulatory and rhythm-generating mechanisms governing the sudomotor and vasoconstrictor outflow in human cutaneous nerves. J Physiol 306, Brengelmann, G.L., Johnson, J.M., Hermansen, L. & Rowell, L.B Altered control of skin blood flow during exercise at high internal temperatures. J Appl Physiol 43, Brothers, R.M., Bhella, P.S., Shibata, S., Wingo, J.E., Levine, B.D. & Crandall, C.G. 2009a. Cardiac systolic and diastolic function during whole body heat stress. Am J Physiol Heart Circ Physiol 296, H1150 H1156. Brothers, R.M., Wingo, J.E., Hubing, K.A. & Crandall, C.G. 2009b. The effects of reduced end-tidal carbon dioxide tension on cerebral blood flow during heat stress. J Physiol 587, Brothers, R.M., Zhang, R., Wingo, J.E., Hubing, K.A. & Crandall, C.G. 2009c. Effects of heat stress on dynamic cerebral autoregulation during large fluctuations in arterial blood pressure. J Appl Physiol 107, Buchthal, F., Honcke, P. & Lindhard, J Temperature measurements in human muscle in situ at rest and during muscular work. Acta Physiol Scand 8, Calbet, J.A., Gonzalez-Alonso, J., Helge, J.W., Sondergaard, H., Munch-Andersen, T., Boushel, R. & Saltin, B Cardiac output and leg and arm blood flow during incremental exercise to exhaustion on the cycle ergometer. J Appl Physiol 103, Christensen, E.H., Nielsen, M. & Hannisdahl, B Investigations of the circulation in the skin at the beginning of muscular work. Acta Physiol Scand 4, Crandall, C.G Carotid baroreflex responsiveness in heat-stressed humans. Am J Physiol Heart Circ Physiol 279, H1955 H1962. Crandall, C.G., Farr, D. & Etzel, R.A. 1999a. Cardiopulmonary baroreceptor control of muscle sympathetic nerve activity in heat-stressed humans. Am J Physiol 277(6 Pt 2), H2348 H2352. Crandall, C.G., Levine, B.D. & Etzel, R.A. 1999b. Effect of increasing central venous pressure during passive heating on skin blood flow. J Appl Physiol 86, Crandall, C.G., Zhang, R. & Levine, B.D Effects of whole body heating on dynamic baroreflex regulation of heart rate in humans. Am J Physiol Heart Circ Physiol 279, H2486 H2492. Crandall, C.G., Cui, J. & Wilson, T.E Effects of heat stress on baroreflex function in humans. Acta Physiol Scand 177, Crandall, C.G., Wilson, T.E., Marving, J., Vogelsang, T.W., Kjaer, A., Hesse, B. & Secher, N.H Effects of passive heating on central blood volume and ventricular dimensions in humans. J Physiol 586, Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 419

71 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, Cui, J., Wilson, T.E. & Crandall, C.G. 2002a. Baroreflex modulation of muscle sympathetic nerve activity during cold pressor test in humans. Am J Physiol Heart Circ Physiol 282, H1717 H1723. Cui, J., Wilson, T.E. & Crandall, C.G. 2002b. Baroreflex modulation of sympathetic nerve activity to muscle in heatstressed humans. Am J Physiol Regul Integr Comp Physiol 282, R252 R258. Cui, J., Wilson, T.E. & Crandall, C.G. 2002c. Phenylephrineinduced elevation in arterial blood pressure is attenuated in heat-stressed humans. Am J Physiol Regul Integr Comp Physiol 283, R1221 R1226. Cui, J., Wilson, T.E. & Crandall, C.G. 2004a. Muscle sympathetic nerve activity during lower body negative pressure is accentuated in heat-stressed humans. J Appl Physiol 96, Cui, J., Wilson, T.E. & Crandall, C.G. 2004b. Orthostatic challenge does not alter skin sympathetic nerve activity in heat-stressed humans. Auton Neurosci 116, Cui, J., Zhang, R., Wilson, T.E. & Crandall, C.G. 2004c. Spectral analysis of muscle sympathetic nerve activity in heat-stressed humans. Am J Physiol Heart Circ Physiol 286, H1101 H1106. Damato, A.M., Lau, S.H., Stein, E., Haft, J.I., Kosowsky, B. & Cohen, S.I Cardiovascular response to acute thermal stress (hot dry environment) in unacclimatized normal subjects. Am Heart J 76, Delp, M.D., Armstrong, R.B., Godfrey, D.A., Laughlin, M.H., Ross, C.D. & Wilkerson, M.K Exercise increases blood flow to locomotor, vestibular, cardiorespiratory and visual regions of the brain in miniature swine. J Physiol 533, Doering, T.J., Aaslid, R., Steuernagel, B., Brix, J., Niederstadt, C., Breull, A., Schneider, B. & Fischer, G.C Cerebral autoregulation during whole-body hypothermia and hyperthermia stimulus. Am J Phys Med Rehabil 78, Edvinsson, L Neurogenic mechanisms in the cerebrovascular bed. Autonomic nerves, amine receptors and their effects on cerebral blood flow. Acta Physiol Scand Suppl 427, Eisalo, A Effects of the Finnish sauna on circulation. Ann Med Exptl Biol Fenniae 4(Suppl.), Fan, J.L., Cotter, J.D., Lucas, R.A., Thomas, K., Wilson, L. & Ainslie, P.N Human cardiorespiratory and cerebrovascular function during severe passive hyperthermia: effects of mild hypohydration. J Appl Physiol 105, Frank, O Zur dynamik des hermuskels. Z Biol 32, Frayser, R., Ross, J.C., Levin, H.S., Messer, J.V. & Pines, J Effects of increased environmental temperature on pulmonary diffusing capacity. J Appl Physiol 21, Fritzsche, R.G., Switzer, T.W., Hodgkinson, B.J. & Coyle, E.F Stroke volume decline during prolonged exercise is influenced by the increase in heart rate. J Appl Physiol 86, Fujii, N., Honda, Y., Hayashi, K., Soya, H., Kondo, N. & Nishiyasu, T Comparison of hyperthermic hyperpnea elicited during rest and submaximal, moderate-intensity exercise. J Appl Physiol 104, Gebber, G.L. & Snyder, D.W Hypothalamic control of baroreceptor reflexes. Am J Physiol 218, Glaser, E.M., Berridge, F.R. & Prior, K.M Effects of heat and cold on the distribution of blood within the human body. Clin Sci 9, González-Alonso, J. & Calbet, J.A Reductions in systemic and skeletal muscle blood flow and oxygen delivery limit maximal aerobic capacity in humans. Circulation 107, González-Alonso, J., Mora-Rodríguez, R., Bellow, P.R. & Coyle, E.F Dehydration reduces cardiac output and increases systemic and cutaneous vascular resistance during exercise. J Appl Physiol 79, González-Alonso, J., Mora-Rodriguez, R., Below, P.R. & Coyle, E.F Dehydration markedly impairs cardiovascular function in hyperthermic endurance athletes during exercise. J Appl Physiol 82, González-Alonso, J., Calbet, J.A. & Nielsen, B Muscle blood flow is reduced with dehydration during prolonged exercise in humans. J Physiol 513(Pt 3), González-Alonso, J., Calbet, J.A. & Nielsen, B. 1999a. Metabolic and thermodynamic responses to dehydration-induced reductions in muscle blood flow in exercising humans. J Physiol 520(Pt 2), González-Alonso, J., Mora-Rodriguez, R. & Coyle, E.F. 1999b. Supine exercise restores arterial blood pressure and skin blood flow despite dehydration and hyperthermia. Am J Physiol 277, H576 H583. González-Alonso, J., Teller, C., Andersen, S.L., Jensen, F.B., Hyldig, T. & Nielsen, B. 1999c. Influence of body temperature on the development of fatigue during prolonged exercise in the heat. J Appl Physiol 86, González-Alonso, J., Mora-Rodriguez, R. & Coyle, E.F. 2000a. Stroke volume during exercise: interaction of environment and hydration. Am J Physiol Heart Circ Physiol 278, H321 H330. González-Alonso, J., Quistorff, B., Krustrup, P., Bangsbo, J. & Saltin, B. 2000b. Heat production in human skeletal muscle at the onset of intense dynamic exercise. J Physiol 524(Pt 2), González-Alonso, J., Dalsgaard, M.K., Osada, T., Volianitis, S., Dawson, E.A., Yoshiga, C.C. & Secher, N.H Brain and central haemodynamics and oxygenation during maximal exercise in humans. J Physiol 557, González-Alonso, J., Crandall, C.G. & Johnson, J.M The cardiovascular challenge of exercising in the heat. J Physiol 586, Hales, J.R.S., Hubbard, R.W. & Gaffin, S.L Limitation of heat tolerance. In: C. Blatteis & M. Fregly (eds) Handbook of Physiology Environmental Physiology, pp American Physiological Society, Bethesda, MD. Hanada, A., Sander, M. & Gonzalez-Alonso, J Human skeletal muscle sympathetic nerve activity, heart rate and limb haemodynamics with reduced blood oxygenation and exercise. J Physiol 551, Heistad, D. & Kontos, H Cerebral circulation. In: J. Shepherd & F. Abboud (eds) Handbook of Physiology, Section 2: The Cardiovascular System, pp American Physiological Society, Bethesda, MD. 420 Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

72 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system Higginbotham, M.B., Morris, K.G., Williams, R.S., McHale, P.A., Coleman, R.E. & Cobb, F.R Regulation of stroke volume during submaximal and maximal upright exercise in normal man. Circ Res 58, Ho, C.W., Beard, J.L., Farrell, P.A., Minson, C.T. & Kenney, W.L Age, fitness, and regional blood flow during exercise in the heat. J Appl Physiol 82, Ichinose, M., Saito, M., Fujii, N., Ogawa, T., Hayashi, K., Kondo, N. & Nishiyasu, T Modulation of the control of muscle sympathetic nerve activity during incremental leg cycling. J Physiol 586, Ide, K., Boushel, R., Sorensen, H.M., Fernandes, A., Cai, Y., Pott, F. & Secher, N.H Middle cerebral artery blood velocity during exercise with beta-1 adrenergic and unilateral stellate ganglion blockade in humans. Acta Physiol Scand 170, Johnson, J.M Exercise and the cutaneous circulation. Exer Sports Sci Rev 20, Johnson, J.M. & Proppe, D.W Cardiovascular adjustments to heat stress. In: C. Blatteis & M. Fregly (eds) Handbook of Physiology: Adaptations to the Environment, pp American Physiological Society, Bethesda, MD. Johnson, J.M. & Rowell, L.B Forearm skin and muscle vascular responses to prolonged leg exercise in man. J Appl Physiol 39, Katz, A.M Physiology of the Heart. Lippincott William & Wilkins, Philadelphia. Keller, D.M., Cui, J., Davis, S.L., Low, D.A. & Crandall, C.G Heat stress enhances arterial baroreflex control of muscle sympathetic nerve activity via increased sensitivity of burst gating, not burst area, in humans. J Physiol 573, Keller, D.M., Sander, M. & Crandall, C.G The effect of passive leg heating on alpha-adrenergic mediated vasoconstriction in humans. Faseb J 21, Abstract Keller, D., Low, D.A., Wingto, J., Brothers, R.M., Hastings, J., Davis, S.L. & Crandall, C.G Acute volume expansion preserves orthostatic tolerance during whole-body heat stress in humans. J Physiol 587, Kenney, W.L. & Johnson, J.M Control of skin blood flow during exercise. Med Sci Sports Exerc 24, Koroxenidis, G.T., Shepherd, J.T. & Marshall, R.J Cardiovascular response to acute heat stress. J Appl Physiol 16, Laughlin, M.H. & Armstrong, R.B Rat muscle blood flows as a function of time during prolonged slow treadmill exercise. Am J Physiol 244, H814 H824. Lee, K., Jackson, D.N., Cordero, D.L., Nishiyasu, T., Peters, J.K. & Mack, G.W Change in spontaneous baroreflex control of pulse interval during heat stress in humans. J Appl Physiol 95, van Lieshout, J.J. & Secher, N.H Point:Counterpoint: Sympathetic activity does/does not influence cerebral blood flow. Point: Sympathetic activity does influence cerebral blood flow. J Appl Physiol 105, Lind, A.R., Leithead, C.S. & McNicol, G.W Cardiovascular changes during syncope induced by tilting men in the heat. J Appl Physiol 25, Lindhard, J Investigations into the conditions governing the temperature of the body. In: L. Mylius-Erichsen (ed.) Danmark-ekspeditionen til Gronlands Nordostkyst , pp Bianco Lunos Bogtrykkeri, Kobenhavn. Low, D.A., Wingo, J.E., Keller, D.M., Davis, S.L., Zhang, R. & Crandall, C.G Cerebrovascular responsiveness to steady-state changes in end-tidal CO 2 during passive heat stress. J Appl Physiol 104, Low, D.A., Wingo, J.E., Keller, D.M., Davis, S.L., Cui, J., Zhang, R. & Crandall, C.G Dynamic cerebral autoregulation during passive heat stress in humans. Am J Physiol Regul Integr Comp Physiol 296, R1598 R1605. McKirnan, M.D., Gray, C.G. & White, F.C Plateau in muscle blood flow during prolonged exercise in miniature swine. J Appl Physiol 66, Minson, C.T., Wladkowski, S.L., Cardell, A.F., Pawelczyk, J.A. & Kenney, W.L Age alters cardiovascular response to direct passive heating. J Appl Physiol 84, Montain, S.J. & Coyle, E.F. 1992a. Fluid ingestion during exercise increases skin blood flow independent of increases in blood volume. J Appl Physiol 73, Montain, S.J. & Coyle, E.F. 1992b. Influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J Appl Physiol 73, Montain, S.J., Sawka, M.N., Latzka, W.A. & Valeri, C.R Thermal and cardiovascular strain from hypohydration: influence of exercise intensity. Int J Sports Med 19, Mortensen, S.P., Dawson, E.A., Yoshiga, C.C., Dalsgaard, M.K., Damsgaard, R., Secher, N.H. & Gonzalez-Alonso, J Limitations to systemic and locomotor limb muscle oxygen delivery and uptake during maximal exercise in humans. J Physiol 566, Mortensen, S.P., Damsgaard, R., Dawson, E.A., Secher, N.H. & Gonzalez-Alonso, J Restrictions in systemic and locomotor skeletal muscle perfusion, oxygen supply and VO 2 during high-intensity whole-body exercise in humans. J Physiol 586, Müller, O Uber die Blutverteilung im menschichen Korper unter dem Einfluss thermischer Reize. Arch Klin Med 82, Nadel, E.R., Fortney, S.M. & Wenger, C.B Effect of hydration state on circulatory and thermal regulations. J Appl Physiol 49, Nelson, E. & Rennels, M Innervation of intracranial arteries. Brain 93, Nielsen, M Die Regulation der Korpertemperature bei Muskelarbeit. Skand Arch Physiol 79, Nielsen, B. & Nielsen, M Body temperature during work at different environmental temperatures. Acta Physiol Scand 56, Nielsen, B. & Nielsen, M. 1965a. Influence of passive and active heating on the temperature regulation of man. Acta Physiol Scand 64, Nielsen, B. & Nielsen, M. 1965b. On the regulation of sweat secretion in exercise. Acta Physiol Scand 64, Nielsen, B., Savard, G., Richter, E.A., Hargreaves, M. & Saltin, B Muscle blood flow and muscle metabolism Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 421

73 Heat stress and the cardiovascular system Æ C G Crandall and J González-Alonso Acta Physiol 2010, 199, during exercise and heat stress. J Appl Physiol 69, Nielsen, B., Hales, J.R., Strange, S., Christensen, N.J., Warberg, J. & Saltin, B Human circulatory and thermoregulatory adaptations with heat acclimation and exercise in a hot, dry environment. J Physiol 460, Nielsen, B., Strange, S., Christensen, N.J., Warberg, J. & Saltin, B Acute and adaptive responses in humans to exercise in a warm, humid environment. Pflugers Arch 434, Niimi, Y., Matsukawa, T., Sugiyama, Y., Shamsuzzaman, A.S.M., Ito, H., Sobue, G. & Mano, T Effect of heat stress on muscle sympathetic nerve activity in humans. J Auton Nerv Syst 63, Nybo, L. & Nielsen, B Middle cerebral artery blood velocity is reduced with hyperthermia during prolonged exercise in humans. J Physiol 534, Nybo, L., Jensen, T., Nielsen, B. & Gonzalez-Alonso, J Effects of marked hyperthermia with and without dehydration on VO 2 kinetics during intense exercise. J Appl Physiol 90, Nybo, L., Moller, K., Volianitis, S., Nielsen, B. & Secher, N.H Effects of hyperthermia on cerebral blood flow and metabolism during prolonged exercise in humans. J Appl Physiol 93, Ogoh, S., Brothers, R.M., Eubank, W.L. & Raven, P.B Autonomic neural control of the cerebral vasculature: acute hypotension. Stroke 39, Parker, J.O. & Case, R.B Normal left ventricular function. Circulation 60, Patterson, S.W. & Starling, E.H On the mechanical factors which determine the output of the ventricles. J Physiol (Lond) 48, Patterson, S.W., Piper, H. & Starling, E.H The regulation of the ventricles. J Physiol (Lond) 48, Paulson, O.B., Strandgaard, S. & Edvinsson, L Cerebral autoregulation. Cerebrovasc Brain Metab Rev 2, Peters, J.K., Nishiyasu, T. & Mack, G.W Reflex control of the cutaneous circulation during passive body core heating in humans. J Appl Physiol 88, Poliner, L.R., Dehmer, G.J., Lewis, S.E., Parkey, R.W., Blomqvist, C.G. & Willerson, J.T Left ventricular performance in normal subjects: a comparison of the responses to exercise in the upright and supine positions. Circulation 62, Rasmussen, P., Stie, H., Nielsen, B. & Nybo, L Enhanced cerebral CO 2 reactivity during strenuous exercise in man. Eur J Appl Physiol 96, Ray, C.A. & Gracey, K.H Augmentation of exerciseinduced muscle sympathetic nerve activity during muscle heating. J Appl Physiol 82, Reis, D.J. & Cuénod, M Central neural regulation of carotid baroreceptor reflexes in the cat. Am J Physiol 209, Remensnyder, J.P., Mitchell, J.H. & Sarnoff, S.J Functional sympatholysis during muscular activity. Observations on influence of carotid sinus on oxygen uptake. Circ Res 11, Rosenmeier, J.B., Hansen, J. & Gonzalez-Alonso, J Circulating ATP-induced vasodilatation overrides sympathetic vasoconstrictor activity in human skeletal muscle. J Physiol 558, Rowell, L.B Human cardiovascular adjustments to exercise and thermal stress. Physiol Rev 54, Rowell, L.B Cardiovascular adjustments to thermal stress. In: L.B. Rowell (ed.) Handbook of Physiology, The Cardiovascular System, pp American Physiological Society, Bethesda, MD. Rowell, L.B. 1986a. Circulatory adjustments to dynamic exercise. In: L.B. Rowell (ed.) Human Circulation Regulation During Physical Stress, pp Oxford University Press, New York. Rowell, L.B. 1986b. Circulatory adjustments to dynamic exercise and heat stress: competing controls. In: L.B. Rowell (ed.) Human Circulation Regulation During Physical Stress, pp Oxford, New York. Rowell, L.B. 1986c. Thermal stress. In: L.B. Rowell (ed.) Human Circulation Regulation During Physical Stress, pp Oxford University Press, New York. Rowell, L.B Hyperthermia: a hyperadrenergic state. Hypertension 15, Rowell, L.B Ideas about control of skeletal and cardiac muscle blood flow ( ): cycles of revision and new vision. J Appl Physiol 97, Rowell, L.B., Blackmon, J.R., Martin, R.H., Mazzarella, J.A. & Bruce, R.A Hepatic clearance of indocyanine green in man under thermal and exercise stresses. J Appl Physiol 20, Rowell, L.B., Marx, H.J., Bruce, R.A., Conn, R.D. & Kusumi, F Reductions in cardiac output, central blood volume, and stroke volume with thermal stress in normal men during exercise. J Clin Invest 45, Rowell, L.B., Kraning, K.K., 2nd, Kennedy, J.W. & Evans, T.O Central circulatory responses to work in dry heat before and after acclimatization. J Appl Physiol 22, Rowell, L.B., Brengelmann, G.L. & Murray, J.A. 1969a. Cardiovascular responses to sustained high skin temperature in resting man. J Appl Physiol 27, Rowell, L.B., Murray, J.A., Brengelmann, G.L. & Kraning, K.K., 2nd. 1969b. Human cardiovascular adjustments to rapid changes in skin temperature during exercise. Circ Res 24, Rowell, L.B., Brengelmann, G.L., Blackmon, J.R. & Murray, J.A Redistribution of blood flow during sustained high skin temperature in resting man. J Appl Physiol 28, Rowell, L.B., Detry, J.R., Profant, G.R. & Wyss, C Spanchnic vasoconstriction in hyperthermic man role of falling blood pressure. J Appl Physiol 31, Saltin, B. & Stenberg, J Circulatory response to prolonged severe exercise. J Appl Physiol 19, Sarnoff, S.J. & Mitchell, J.H The regulation of the performance of the heart. Am J Med 30, Savard, G.K., Nielsen, B., Laszczynska, J., Larsen, B.E. & Saltin, B Muscle blood flow is not reduced in humans during moderate exercise and heat stress. J Appl Physiol 64, Sawka, M.N. & Coyle, E.F Influence of body water and blood volume on thermoregulation and exercise performance in the heat. Exerc Sport Sci Rev 27, Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x

74 Acta Physiol 2010, 199, C G Crandall and J González-Alonso Æ Heat stress and the cardiovascular system Sawka, M.N., Knowlton, R.G. & Critz, J.B Thermal and circulatory responses to repeated bouts of prolonged running. Med Sci Sports 11, Semenza, J.C., Rubin, C.H., Falter, K.H., Selanikio, J.D., Flanders, W.D., Howe, H.L. & Wilhelm, J.L Heatrelated deaths during the July 1995 heat wave in Chicago. N Engl J Med 335, Semenza, J.C., McCullough, J.E., Flanders, W.D., McGeehin, M.A. & Lumpkin, J.R Excess hospital admissions during the July 1995 heat wave in Chicago [see comments]. Am J Prev Med 16, Strandgaard, S. & Sigurdsson, S.T Point:Counterpoint: Sympathetic activity does/does not influence cerebral blood flow. Counterpoint: Sympathetic nerve activity does not influence cerebral blood flow. J Appl Physiol 105, discussion Strom, G Central nervous regulation of body temperature. In: J. Field, H.W. Magoun, V.E. Hall (eds) Handbook of Physiology, Section 1: Neurophysiology, Vol. II, pp Waverly Press, Washington, DC. Templeton, G.H., Ecker, R.R. & Mitchell, J.H Left ventricular stiffness during diastole and systole: the influence of changes in volume and inotropic state. Cardiovasc Res 6, Umeyama, T., Kugimiya, T., Ogawa, T., Kandori, Y., Ishizuka, A. & Hanaoka, K Changes in cerebral blood flow estimated after stellate ganglion block by single photon emission computed tomography. J Auton Nerv Syst 50, Van Lieshout, J.J., Wieling, W., Karemaker, J.M. & Secher, N.H Syncope, cerebral perfusion, and oxygenation. J Appl Physiol 94, Vanhoutte, P.M., Verbeuren, T.J. & Webb, R.C Local modulation of adrenergic neuroeffector interaction in the blood vessel wall. Physiol Rev 61, Visocchi, M., Chiappini, F., Cioni, B. & Meglio, M Cerebral blood flow velocities and trigeminal ganglion stimulation. A transcranial Doppler study. Stereotact Funct Neurosurg 66, Vogiatzis, I., Athanasopoulos, D., Habazettl, H., Kuebler, W., Wagner, H., Roussos, C., Wagner, P. & Zakynthinos, S Intercostal muscle blood flow limitation in athletes during maximal exercise. J Physiol 587, Wilson, T.E., Cui, J. & Crandall, C.G Absence of arterial baroreflex modulation of skin sympathetic activity and sweat rate during whole-body heating in humans. J Physiol 536, Wilson, T.E., Cui, J. & Crandall, C.G. 2002a. Effect of wholebody and local heating on cutaneous vasoconstrictor responses in humans. Auton Neurosci 97, Wilson, T.E., Cui, J., Zhang, R., Witkowski, S. & Crandall, C.G. 2002b. Skin cooling maintains cerebral blood flow velocity and orthostatic tolerance during tilting in heated humans. J Appl Physiol 93, Wilson, T.E., Cui, J., Zhang, R. & Crandall, C.G Heat stress reduces cerebral blood velocity and markedly impairs orthostatic tolerance in humans. Am J Physiol Regul Integr Comp Physiol 291, R1443 R1448. Wilson, T.E., Tollund, C., Yoshiga, C.C., Dawson, E.A., Nissen, P., Secher, N.H. & Crandall, C.G Effects of heat and cold stress on central vascular pressure relationships during orthostasis in humans. J Physiol 585, Wilson, T.E., Brothers, R.M., Tollund, C., Dawson, E., Nissen, P., Yoshiga, C., Jons, C., Secher, N. & Crandall, C.G Effect of thermal stress on Frank Starling relations in humans. J Physiol 587, Yamazaki, F. & Sone, R Modulation of arterial baroreflex control of heart rate by skin cooling and heating in humans. J Appl Physiol 88, Yamazaki, F., Matsumura, N., Nagata, J., Ando, A. & Imura, T Spontaneous arterial baroreflex control of the heart rate during head-down tilt in heat-stressed humans. Eur J Appl Physiol 85, Yamazaki, F., Yamauchi, K., Tsutsui, Y., Endo, Y., Sagawa, S. & Shiraki, K Whole body heating reduces the baroreflex response of sympathetic nerve activity during Valsalva straining. Auton Neurosci 103, Zhang, R., Zuckerman, J.H., Iwasaki, K., Wilson, T.E., Crandall, C.G. & Levine, B.D Autonomic neural control of dynamic cerebral autoregulation in humans. Circulation 106, Ó 2010 The Authors Journal compilation Ó 2010 Scandinavian Physiological Society, doi: /j x 423

75 Johansson et al. Critical Care (2017) 21:25 DOI /s REVIEW Shock induced endotheliopathy (SHINE) in acute critical illness - a unifying pathophysiologic mechanism PärIngemar Johansson 1,2,3*, Jakob Stensballe 1,4 and SisseRye Ostrowski 1 Open Access Abstract One quarter of patients suffering from acute critical illness such as severe trauma, sepsis, myocardial infarction (MI) or post cardiac arrest syndrome (PCAS) develop severe hemostatic aberrations and coagulopathy, which are associated with excess mortality. Despite the different types of injurious hit, acutely critically ill patients share several phenotypic features that may be driven by the shock. This response, mounted by the body to various life-threatening conditions, is relatively homogenous and most likely evolutionarily adapted. We propose that shock-induced sympatho-adrenal hyperactivation is a critical driver of endothelial cell and glycocalyx damage (endotheliopathy) in acute critical illness, with the overall aim of ensuring organ perfusion through an injured microvasculature. We have investigated more than 3000 patients suffering from different types of acute critical illness (severe trauma, sepsis, MI and PCAS) and have found a potential unifying pathologic link between sympathoadrenal hyperactivation, endotheliopathy, and poor outcome. We entitled this proposed disease entity, shockinduced endotheliopathy (SHINE). Here we review the literature and discuss the pathophysiology of SHINE. Background Acute critical illness such as trauma, sepsis, myocardial infarction (MI) and post cardiac arrest syndrome (PCAS) affects more than five million patients in the EU annually [1]. Approximately one quarter of acutely critically ill patients develop severe hemostatic aberrations resulting in coagulopathy [2 4], which in patients suffering from severe injury is entitled trauma-induced coagulopathy (TIC) [4, 5], and in patients with sepsis and PCAS (and by some also in trauma [6]) entitled disseminated intravascular coagulation (DIC) [7 10]. Acutely critically ill patients with coagulopathy have been reported to have three to four times higher mortality rates than their counterparts without coagulopathy, translating into a mortality rate of approximately 50%, which has remained virtually constant for decades [4, 7, 10]. In studies of trauma patients, increasing injury severity score (ISS) is associated with progressive hypocoagulability * Correspondence: per.johansson@regionh.dk 1 Capital Region Blood Bank, Rigshospitalet Section for Transfusion Medicine, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej, 9DK-2100 Copenhagen, Denmark 2 Department of Surgery, University of Texas Health Medical School, Houston, TX, USA Full list of author information is available at the end of the article [11, 12]. This could be regarded as counterintuitive from an evolutionary perspective, as these patients are at high risk of exsanguination and, therefore, would need an intact or even improved hemostatic capacity of blood flow. We have proposed that the coagulopathy observed in these patients is a compensatory mechanism counterbalancing the shockinduced pro-thrombotic vascular endothelium in the microcirculation in order to secure sufficient organ perfusion in conditions with shock [12, 13]. Importantly, systemic endothelial injury seems pivotal for the development of organ failure and ensuing poor outcome [14, 15], pointing to a possible explanation of the association between coagulopathy and poor outcome in acute critical illness [8, 10, 16, 17]. The endothelium is one of the largest organs in the body, with a total weight of approximately 1 kg and a surface area of approximately 5000 m 2 [18]. Endothelial cells form the innermost lining of all blood and lymphatic vessels and extend to all reaches of the vertebrate body. Far from being an inert layer of nucleated cellophane, the endothelium partakes in a wide array of physiological functions, including control of vasomotor tone, maintenance of blood fluidity, regulated transfer of The Author(s) Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

76 Johansson et al. Critical Care (2017) 21:25 Page 2 of 7 water, nutrients and leukocytes across the vascular wall, innate and acquired immunity, angiogenesis and establishment of a unique dialogue between the underlying tissue and the flowing blood [18]. It is also recognized that the endothelium plays a critical role in a multitude of diseases, such as arteriosclerosis, malignancy and acute inflammatory diseases either as a primary determinant of pathophysiology or as a victim of collateral damage [19, 20]. Under normal conditions the endothelium is anticoagulated by a number of natural anticoagulant systems including the negatively charged luminal surface layer, the glycocalyx, which is rich in heparonoids and interacts with antithrombin [21]. Furthermore, tissue factor pathway inhibitor (TFPI) and the protein C/thrombomodulin system also contribute to endothelial anticoagulation along with endothelial release of tissue-type plasminogen activator (tpa) and urokinase-type plasminogen activator (upa) that dissolves forming clots [22]. Hence, we propose that shedding, degradation and/or release of the glycocalyx and the natural anticoagulant and profibrinolytic factors from the injured endothelium induces the profound hypocoagulability observed in acute critically ill patients with shock [12]. In trauma patients, TIC is present already at the scene of the accident in the most severely injured, shocked patients [23] indicating a potential contribution of the sympatho-adrenal system to this early coagulopathy. Cannon described in 1915 how the hormone adrenaline, released immediately upon severe stress, mobilizes an emergency response denoted the fight or flight response, and furthermore that the sympatho-adrenal activation orchestrates changes in blood supply, sugar availability and the blood s clotting capacity in a marshalling of resources keyed to the violent display of energy [24]. We propose that the shock-induced sympatho-adrenal hyperactivation and ensuing excessive increase in circulating levels of catecholamines, not only activates but also directly inflicts systemic damage to the endothelium, including the microcirculation [25, 26]. Apart from the obvious increased risk of microvascular occlusion secondary to pro-thrombotic microcirculation in these patients, capillary leakage also significantly contributes to disease progression due to hypovolemia, edema, tissue hypoxia and exacerbated shock, resulting in a viscous circle with sustained sympatho-adrenal hyperactivation and release of large amounts of catecholamines, further compromising the microvasculature [27] (Fig. 1). Here we describe and discuss the pathophysiology of shock-induced endotheliopathy (SHINE), a proposed new disease entity with unifying pathological change observed in acutely critically ill patients challenged by shock. Shock-induced endotheliopathy (SHINE) We propose that shock, and its effect on the sympathoadrenal system, the endothelium, including the glycocalyx and the hemostatic cells in the circulating blood results in phenotypic features that characterize the clinical condition of patients suffering acute critical illness, despite the different types of injurious hit they suffer [6, 9, 15, 27 30]. The catecholamine-induced damage to the endothelium is responsible for endothelial breakdown resulting in glycocalyx shedding, breakdown of tight junctions with capillary leakage and a procoagulant microvasculature that further reduces oxygen delivery due to increased tissue pressure and microvascular thrombosis creating a vicious circle that ultimately results in organ failure. The early genetic responses to severe trauma, burn injury and endotoxemia are similar [31], indicating that the response mounted by the body to various acute critical conditions accompanied by shock, is relatively homogenous and most likely evolutionarily adapted [12]. Endotheliopathy of traumatic shock We have investigated the degree of coagulopathy, sympatho-adrenal activation (plasma catecholamines) and endothelial injury (circulating biomarkers of endothelial cell (soluble thrombomodulin (stm)) and glycocalyx (syndecan-1) damage) in three independent cohorts of severely injured patients (total number 579) [5, 16, 32 35]. Here we found strong and independent associations between high injury severity, high plasma adrenaline Fig. 1 Shock-induced endotheliopathy (SHINE). Schematic illustration of the changes in the vascular compartment with increasing disease severity and increasing sympatho-adrenal activation (Original figure)

77 Johansson et al. Critical Care (2017) 21:25 Page 3 of 7 level, profound hypocoagulability and high circulating syndecan-1 and stm levels. High plasma adrenaline was a strong and independent predictor of increased mortality [32] and hypocoagulability [36] and, importantly, despite comparable injury severity, trauma patients with the highest syndecan-1 levels (reflecting the highest degree of glycocalyx damage) had several-fold higher mortality [16, 33]. This emphasizes the pivotal importance of the state of the endothelium for outcome in these patients and also points towards a possible genetic predisposition of the endothelial response to shock. Furthermore, we found a significantly different sympatho-adrenal and endothelial response to the injurious hit in older vs. younger trauma patients, indicating that patient age also appears to significantly influence the response that is mounted, including the degree of endotheliopathy [37]. This is in accordance with the well-described association between higher age and progressive disruption and dysfunction of the endothelium, with the most profound endothelial disruption observed in smokers and patients with diabetes, hypertension or atherosclerosis [20, 38]. In addition to age, gender also significantly influences the endogenous trauma shock response [39] and both age and male gender are strong and independent predictors of multiple organ failure, an outcome closely linked to endotheliopathy, following severe trauma [40]. The critical importance of glycocalyx shedding in TIC was further illustrated by our finding that the most severely injured trauma patients displayed evidence of endogenous heparinization, as evaluated by whole blood thrombelastography (TEG) [35]. Endogenous heparinization is the result of the shedding of the glycocalyx, including heparan sulphate having the same functional effects as heparin on the hemostatic system. Also, damage to the endothelial cells induces release of thrombomodulin in its soluble form, which retains its anticoagulant effects also when circulating in the blood. Patients with evident endogenous heparinization displayed four-fold higher plasma syndecan-1 levels, strongly indicating that release of heparin-like constituents from the glycocalyx induced the endogenous heparinization. Patients with endogenous heparinization also had higher transfusion requirements, higher stm levels and lower protein C levels compared to patients without endogenous heparinization. This emphasizes that the endotheliopathy included both extensive endothelial cell and glycocalyx damage [35, 41 44]. It should be noted, however, that these intriguing data are only observations and as such are hypothesis-generating, and currently there is no firm evidence available from RCTs to clarify whether endotheliopathy merely reflects greater disease severity, which in turn is known to relate to more organ dysfunction, or a severity-independent association with organ injury. Endotheliopathy of septic shock Septic coagulopathy evidenced by DIC has for decades been associated with poor outcome [7, 8] and the accompanying endothelial dysfunction and injury are both hallmarks and drivers of the poor outcome [8, 29]. Based on the hypothesis that coagulopathy is a surrogate marker and a result of systemic endotheliopathy, we conducted a study investigating patients (n = 321) with varying degrees of infectious disease ranging from systemic inflammatory response syndrome (SIRS) without infection or with local infection, to sepsis, severe sepsis or septic shock [45]. Here we found that plasma syndecan-1 and stm increased progressively and significantly across groups with increasing infectious severity and correlated significantly with organ failure as measured by the sequential organ failure assessment (SOFA) score in all groups. Furthermore, plasma levels of catecholamines, syndecan-1 and stm were significantly higher in non-survivors compared to survivors and high levels of both catecholamines, syndecan-1 and stm were all independent predictors of excess mortality, linking sympatho-adrenal hyperactivation and endothelial damage to outcome in patients with sepsis. Patients with septic shock per definition receive vasopressor treatment, most often noradrenaline. Given this, it could be speculated whether the high therapeutic noradrenaline concentrations further promote endotheliopathy in these patients. We investigated this in a small study of patients (n = 67) of whom 21% received noradrenaline infusion at the time of blood sampling [46]. The study demonstrated that the levels of a broad range of biomarkers reflecting endothelial damage, including syndecan-1 and stm, did not differ between patients with or without noradrenaline infusion, indicating that endotheliopathy in patients with septic shock was not further aggravated by catecholamine infusion [46]. Similarly, there was a strong association between endotheliopathy and organ failure in a large multicenter study of 1103 critically ill patients predominantly suffering from sepsis [47], demonstrating that patients with sepsis had higher plasma levels of syndecan-1 and stm (more excessive endothelial damage) than non-infected patients. When stratifying the patients into quartiles based on stm levels at study enrollment, mortality could be differentiated across all four quartiles during the entire follow-up period, with the highest mortality in the highest stm quartiles, even after adjusting for other prognostic variables. Importantly, high syndecan-1 and stm levels independently predicted liver and renal failure, respectively, and high stm was further associated with increased risk of development of multiple organ failure. In sensitivity analysis, a composite endpoint of circulatory failure or death was created to overcome potential lead bias, as inotropic/vasopressor drugs are

78 Johansson et al. Critical Care (2017) 21:25 Page 4 of 7 often removed from patients bound to die. After adjusting for confounders, both syndecan-1 and stm study enrollment independently predicted the risk of circulatory failure or death, further pointing towards the central role of endotheliopathy for the pathophysiology related to outcome in patients with septic shock [47]. Finally, in a smaller cohort of 184 patients with severe sepsis or septic shock we found an independent association between high circulating syndecan-1 levels and coagulopathy evaluated by TEG, further linking endotheliopathy and coagulopathy also in sepsis [45]. Though it has been evident for decades that endothelial injury is a hallmark of sepsis [8, 27, 29], new data keep emerging that further reveal the pathophysiology of endothelial cell and glycocalyx damage in sepsis and its association with disease severity, including the applicability of biomarkers for outcome [48 51]. Similar to traumatic endotheliopathy, the findings described here are observational and, hence, no causality can be inferred. Endotheliopathy of cardiogenic shock and cardiac arrest Cardiac arrest is the ultimate ischemia-reperfusion hit to the body. PCAS represents the systemic response to the global ischemia-reperfusion injury [15], which involves profound endothelial injury and ensuing microcirculatory dysfunction and failure secondary to capillary leakage, tissue/organ edema and hypoxia and increased blood cell adhesion to the activated/injured endothelium. The consequence of this global ischemia-reperfusion injury to the endothelium is a sepsis-like inflammatory response [9, 15, 30] that ultimately drives organ failure similarly to that observed in sepsis. In 2007, Rehm and colleagues provided the first evidence in humans for shedding of the endothelial glycocalyx in conditions with ischemia-reperfusion [52]. In three groups of surgical patients (patients undergoing thoracic aortic surgery with deep hypothermic cardiac arrest, patients undergoing cardiac surgery on cardiopulmonary bypass and patients undergoing surgery for an abdominal aortic aneurysm) it was found that global and regional ischemia was followed by an increase in both syndecan-1 and heparan sulfate, two constituents of the endothelial glycocalyx [52], a finding confirmed by later studies [53]. Patients resuscitated from cardiac arrest frequently demonstrate profound hypocoagulability and hyperfibrinolysis of the flowing blood along with shedding of the glycocalyx [54, 55]. In a post-hoc analysis of 163 patients included at our center, Rigshospitalet, in The Targeted Temperature Management at 33 degrees versus 36 degrees after Cardiac Arrest (TTM) trial [56], we found that catecholamines correlated strongly with syndecan-1 and stm plasma levels i.e. biomarkers reflecting endothelial glycocalyx and cell damage [57]. Overall 180-day mortality was 35% and both plasma adrenaline and stm levels were the strongest, and independent, predictors of mortality [57]. This finding is in line with our previous study of 678 patients with acute STelevation myocardial infarction (STEMI), demonstrating that admission levels of plasma adrenaline, syndecan-1 and stm were highly correlated with the highest levels of adrenaline and syndecan-1 in patients with cardiogenic shock [38]. Furthermore, STEMI patients admitted to ICU displayed the highest syndecan-1 plasma levels and high levels of adrenaline, syndecan-1 and stm were strong predictors of poor outcome, including heart failure and mortality [38]. Together these findings indicate that sympathoadrenal hyperactivation and endothelial damage are inter-correlated and strong predictors of mortality in conditions with cardiogenic shock [38, 57], and furthermore that myocardial infarction alone appears also to inflict significant systemic endothelial damage, possibly driven in part by the parallel increase in circulating catecholamines, albeit evidence from prospective randomized trials are lacking [38]. The finding, however, is in alignment with previous studies reporting high circulating levels of glycocalyx components (syndecan-1, heparan sulphate) in patients with cardiogenic shock, with high levels being strong predictors of excess mortality [58]. Discussion In the observational data presented here from more than 3000 patients with different types of acute critical illness including different types of shock, high circulating catecholamine levels are independently associated with endotheliopathy and are predictive of poor outcome (both short-term and long-term mortality) and, furthermore, that this shock-induced endotheliopathy is statistically linked to the development of organ failure and death. Given that shock and endothelial disruption and damage coincide in patients with the most severe form of acute critical illness, a mechanistic link is suggested between sympatho-adrenal hyperactivation and the endothelial phenotype, and that this shock-induced endotheliopathy (SHINE), may be a unifying pathophysiologic mechanism, linked to outcome, albeit this awaits further confirmation [12, 28]. Recently, a link between sympatho-adrenal hyperactivation and endothelial damage was suggested in an animal model of trauma shock demonstrating that both chemical sympathectomy and treatment with β-blockade attenuate endothelial glycocalyx and endothelial cell damage in rats with acute traumatic coagulopathy [59]. This may provide an explanation for the limited success of many large RCTs conducted in acutely critically ill

79 Johansson et al. Critical Care (2017) 21:25 Page 5 of 7 patients in the past decades [60]. Among patients with severe sepsis/septic shock alone, more than 30,000 patients have been enrolled in clinical trials to test anti-coagulant, anti-inflammatory, anti-endotoxin and immune-modulating agents [60, 61]. Yet, not a single agent has convincingly proven to be consistently efficacious and there are still no new drugs on the market with the indication of sepsis, despite tremendous effort worldwide. Similarly, in patients suffering from out of hospital cardiac arrest (OHCA), two small RCTs (77 and 136 patients, respectively) conducted in 2002 reported improved survival in those receiving therapeutic hypothermia targeted at approximately 33 C [62, 63]. However, in a large RCT including 939 patients randomized to temperatures of 33 C or 36 C, there was difference between groups in mortality [56], and a recent meta-analysis of RCTs reported no benefit of mild therapeutic hypothermia on neurologic outcome or mortality in patients who had OHCA [64]. In trauma, mortality has been reduced substantially in the past years as a result of the introduction of damage control surgery and hemostatic resuscitation [65 67]. A recent multicenter RCT in trauma patients with severe hemorrhage demonstrated a significant reduction in early mortality caused by exsanguination, with more aggressive administration of plasma and platelets [68]. Similarly, a recent RCT was prematurely halted due to a significantly increased survival of patients who were resuscitated aggressively based on whole blood TEG compared to conventional coagulation assays [69]. Unfortunately, the excess mortality in patients with TIC has remained unchanged by these improvements, highlighting a therapeutic failure here as well. Given the potential unifying pathologic condition of SHINE across patients with different types of acute critical illness, it could be speculated whether interventions targeting the endothelium and/or the sympatho-adrenal system could be of value here. By 1978, β-blocker therapy had already been reported to have beneficial effects on MI [70] and in a later meta-analysis of RCTs investigating the use of early intravenous beta-blockers in patients with acute coronary syndrome there were significant reductions in the risk of short-term cardiovascular events, including reduction in all-cause mortality [71]. The beneficial effects of β-blocker therapy in these patients have historically been envisioned to be related to reductions in the incidence of arrhythmia and improved cardiac myocyte function. However, we speculate that blockade of the effects of the catecholamine surge on the endothelium, and hereby reduced systemic endotheliopathy, may also have contributed to the improved outcome and this should be investigated further. In a recent small RCT of patients with septic shock and heart rate above 95 beats per minute, 77 patients were randomized to either short-acting β-blocker therapy with Esmolol to maintain heart rate between 80 and 94 beats per minute during their ICU stay or to placebo [72]. Patients receiving β-blocker therapy had lower 28-day mortality compared to the control group (49% vs. 81%, adjusted hazard ratio of 0.39). Taken together these results may indicate that sympathoadrenal hyper-activation may be hazardous for acute critically ill patients and according to our proposed hypothesis, use of β-blocker therapy in these previous trials may have prevented or reduced the catecholamine-induced endotheliopathy, which translated into improved survival in patients suffering from cardiac disease including cardiac arrest, trauma and sepsis. Adequately powered RCTs are necessary to confirm or reject this hypothesis. Conclusion Shock-induced endotheliopathy (SHINE) is observed in acute critical illness and may reflect a potential unifying pathophysiologic mechanism linked to poor outcome. Sympatho-adrenal hyperactivation appears to be a pivotal driver of this condition. Abbreviations DIC: Disseminated intravascular coagulation; MI: Myocardial infarction; OHCA: Out-of-hospital cardiac arrest; PCAS: Post cardiac arrest syndrome; RCT: Randomized controlled trial; SHINE: Shock-induced endotheliopathy; SIRS: Systemic inflammatory response syndrome; SOFA: Sequential organ failure assessment; stm: soluble Thrombomodulin; TEG: Thrombelastography; TFPI: Tissue actor pathway inhibitor; TIC: Trauma-induced coagulopathy; TTM: Targeted temperature management Acknowledgements Not applicable. Funding No funding was provided. Availability of data and materials Not applicable. Authors contributions PJ performed the literature review, wrote the manuscript and reviewed the final version. JS also wrote the manuscript and reviewed the final version. SO also participated in the literature review, wrote the manuscript and reviewed the final version. All authors read and approved the final manuscript. Authors information Not applicable. Competing interests The authors declare that they have no competing interests. Consent for publication Not applicable. Ethics approval and consent to participate Not applicable.

80 Johansson et al. Critical Care (2017) 21:25 Page 6 of 7 Author details 1 Capital Region Blood Bank, Rigshospitalet Section for Transfusion Medicine, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej, 9DK-2100 Copenhagen, Denmark. 2 Department of Surgery, University of Texas Health Medical School, Houston, TX, USA. 3 Centre for Systems Biology, The School of Engineering and Natural Sciences, University of Iceland, Reykjavik, Iceland. 4 Department of Anesthesia, Centre of Head and Orthopedics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark. References 1. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to PLoS Med. 2006;3(11):e Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54(6): Holcomb JB, Minei KM, Scerbo ML, RadwanZA,WadeCE,KozarRA, GillBS,AlbaradoR,McNuttMK,KhanS,etal.Admissionrapid thrombelastography can replace conventional coagulation tests in the emergency department: experience with 1974 consecutive trauma patients. Ann Surg. 2012;256(3): Hess JR, Brohi K, Dutton RP, Hauser CJ, Holcomb JB, Kluger Y, Mackway- Jones K, Parr MJ, Rizoli SB, Yukioka T, et al. The coagulopathy of trauma: a review of mechanisms. J Trauma. 2008;65(4): Johansson PI, Sørensen AM, Perner A, Welling KL, Wanscher M, Larsen CF, Ostrowski SR. Disseminated intravascular coagulation or acute coagulopathy of trauma shock early after trauma? An observational study. Crit Care. 2011;15(6):R Gando S, Sawamura A, Hayakawa M. Trauma, Shock and disseminated intravascular coagulation: lessons from the classical literature. Ann Surg. 2011;254(1): Levi M, Toh CH, Thachil J, Watson HG. Guidelines for the diagnosis and management of disseminated intravascular coagulation. Br J Haematol. 2009;145(1): Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J Med. 2013;369(9): Adrie C, Laurent I, Monchi M, Cariou A, Dhainaou JF, Spaulding C. Post resuscitation disease after cardiac arrest: a sepsis-like syndrome? Curr Opin Crit Care. 2004;10(3): Kim J, Kim K, Lee JH, Jo YH, Kim T, Rhee JE, Kang KW. Prognostic implication of initial coagulopathy in out-of-hospital cardiac arrest. Resuscitation. 2013; 84(1): Johansson PI, Ostrowski SR, Secher NH. Management of major blood loss: an update. Acta Anaesthesiol Scand. 2010;54(9): Johansson PI, Ostrowski SR. Acute coagulopathy of trauma: balancing progressive catecholamine induced endothelial activation and damage by fluid phase anticoagulation. Med Hypotheses. 2010;75: Faust SN, Levin M, Harrison OB, Goldin RD, Lockhart MS, Kondaveeti S, Laszik Z, Esmon CT, Heyderman RS. Dysfunction of endothelial protein C activation in severe meningococcal sepsis. N Engl J Med. 2001;345(6): Levi M, van der PT, Schultz M. Systemic versus localized coagulation activation contributing to organ failure in critically ill patients. Semin Immunopathol. 2012;34(1): doi: /s Neumar RW, Nolan JP, Adrie C, Aibiki M, Berg RA, Bottiger BW, Callaway C, Clark RS, Geocadin RG, Jauch EC, et al. Post-cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. A consensus statement from the International Liaison Committee on Resuscitation (American Heart Association, Australian and New Zealand Council on Resuscitation, European Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Asia, and the Resuscitation Council of Southern Africa); the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiovascular Surgery and Anesthesia; the Council on Cardiopulmonary, Perioperative, and Critical Care; the Council on Clinical Cardiology; and the Stroke Council. Circulation. 2008;118(23): Johansson PI, Henriksen HH, Stensballe J, Gybel-Brask M, Cardenas JC, Baer LA, Cotton BA, Holcomb JB, Wade CE, Ostrowski SR. Traumatic endotheliopathy: a prospective observational study of 424 severely injured patients. Ann Surg. 2016, in press. 17. Cohen MJ, Call M, Nelson M, Calfee CS, Esmon CT, Brohi K. Critical role of activated protein C in early coagulopathy and later organ failure, infection and death in trauma patients. Ann Surg. 2012;255(2): Aird WC. Endothelial cells in health and disease. Florida: Taylor & Francis Group; Hirase T, Node K. Endothelial dysfunction as a cellular mechanism for vascular failure. Am J Physiol Heart Circ Physiol. 2012;302(3):H Broekhuizen LN, Mooij HL, Kastelein JJ, Stroes ES, Vink H, Nieuwdorp M. Endothelial glycocalyx as potential diagnostic and therapeutic target in cardiovascular disease. Curr Opin Lipidol. 2009;20(1): Becker BF, Chappell D, Bruegger D, Annecke T, Jacob M. Therapeutic strategies targeting the endothelial glycocalyx: acute deficits, but great potential. Cardiovasc Res. 2010;87(2): Levi M, van der Poll T. The role of natural anticoagulants in the pathogenesis and management of systemic activation of coagulation and inflammation in critically ill patients. Semin Thromb Hemost. 2008;34(5): Carroll RC, Craft RM, Langdon RJ, Clanton CR, Snider CC, Wellons DD, Dakin PA, Lawson CM, Enderson BL, Kurek SJ. Early evaluation of acute traumatic coagulopathy by thrombelastography. Transl Res. 2009;154(1): Cannon WB. Bodily changes in pain, hunger, fear and rage. An account of recent researches into the function of emotional excitement. 1st ed. New York and London: D. Appleton and Co; Makhmudov RM, Mamedov Y, Dolgov VV, Repin VS. Catecholaminemediated injury to endothelium in rabbit perfused aorta: a quantitative analysis by scanning electron microscopy. Cor Vasa. 1985;27(6): Dolgov VV, Makhmudov RM, Bondarenko MF, Repin VS. Deleterious action of adrenaline on the endothelial lining of the vessels. Arkh Patol. 1984;46(10): Opal SM, van der Poll T. Endothelial barrier dysfunction in septic shock. J Intern Med. 2015;277(3): Holcomb JB. A novel and potentially unifying mechanism for shock induced early coagulopathy. Ann Surg. 2011;254(2): Cohen J, Vincent JL, Adhikari NK, Machado FR, Angus DC, Calandra T, Jaton K, Giulieri S, Delaloye J, Opal S, et al. Sepsis: a roadmap for future research. Lancet Infect Dis. 2015;15(5): Adrie C, Adib-Conquy M, Laurent I, Monchi M, Vinsonneau C, Fitting C, Fraisse F, Dinh-Xuan AT, Carli P, Spaulding C, et al. Successful cardiopulmonary resuscitation after cardiac arrest as a sepsis-like syndrome. Circulation. 2002;106(5): Xiao W, Mindrinos MN, Seok J, Cuschieri J, Cuenca AG, Gao H, Hayden DL, Hennessy L, Moore EE, Minei JP, et al. A genomic storm in critically injured humans. J Exp Med. 2011;208(13): Johansson PI, Stensballe J, Rasmussen LS, Ostrowski SR. High circulating adrenaline levels at admission predict increased mortality after trauma. J Trauma Acute Care Surg. 2012;72(2): Johansson PI, Stensballe J, Rasmussen LS, Ostrowski SR. A high admission syndecan-1 level, a marker of endothelial glycocalyx degradation, is associated with inflammation, protein C depletion, fibrinolysis, and increased mortality in trauma patients. Ann Surg. 2011;254(2): Ostrowski SR, Sorensen AM, Larsen CF, Johansson PI. Thrombelastography and biomarker profiles in acute coagulopathy of trauma: a prospective study. Scand J Trauma Resusc Emerg Med. 2011;19(1): Ostrowski SR, Johansson PI. Endothelial glycocalyx degradation induces endogenous heparinization in patients with severe injury and early traumatic coagulopathy. J Trauma Acute Care Surg. 2012;73(1): Ostrowski SR, Henriksen HH, Stensballe J, Gybel-Brask M, Cardenas JC, Baer LA, Cotton BA, Holcomb JB, Wade CE, Johansson PI: Sympathoadrenal activation and endotheliopathy are drivers of hypocoagulability and hyperfibrinolysis in trauma: A prospective observational study of 404 severely injured patients. J Trauma Acute Care Surg In press. 37. Johansson PI, Sørensen AM, Perner A, Welling KL, Wanscher M, Larsen CF, Ostrowski SR. Elderly trauma patients have high circulating noradrenaline levels but attenuated release of adrenaline, platelets and leukocytes in response to increasing injury severity. Crit Care Med. 2012;40(6): Ostrowski SR, Pedersen SH, Jensen JS, Mogelvang R, Johansson PI. Acute myocardial infarction is associated with endothelial glycocalyx and cell damage and a parallel increase in circulating catecholamines. Crit Care. 2013;17(1):R Schreiber MA, Differding J, Thorborg P, Mayberry JC, Mullins RJ. Hypercoagulability is most prevalent early after injury and in female patients. J Trauma. 2005;58(3):

81 Johansson et al. Critical Care (2017) 21:25 Page 7 of Frohlich M, Lefering R, Probst C, Paffrath T, Schneider MM, Maegele M, Sakka SG, Bouillon B, Wafaisade A. Epidemiology and risk factors of multipleorgan failure after multiple trauma: an analysis of 31,154 patients from the TraumaRegister DGU. J Trauma Acute Care Surg. 2014;76(4): Haywood-Watson R, Pati S, Kozar R, Faz J, Holcomb JB, Gonzalez E. Human micro-vascular barrier disruption after hemorrhagic shock. J Surg Res. 2010;158: Haywood-Watson RJ, Holcomb JB, Gonzalez EA, Peng Z, Pati S, Park PW,WangW,ZaskeAM,MengeT,KozarRA.Modulationofsyndecan- 1 shedding after hemorrhagic shock and resuscitation. PLoS One. 2011;6(8):e Rahbar E, Cardenas JC, Baimukanova G, Usadi B, Bruhn R, Pati S, Ostrowski SR, Johansson PI, Holcomb JB, Wade CE. Endothelial glycocalyx shedding and vascular permeability in severely injured trauma patients. J Transl Med. 2015;13:117. doi: /s : Di Battista AP, Rizoli SB, Lejnieks B, Min A, Shiu MY, Peng HT, Baker AJ, Hutchison MG, Churchill N, Inaba K, et al. Sympathoadrenal activation is associated with acute traumatic coagulopathy and endotheliopathy in isolated brain injury. Shock. 2016;46: Ostrowski SR, Haase N, Müller RB, Moller MH, Pott FC, Perner A, Johansson PI. Association between biomarkers of endothelial injury and hypocoagulability in patients with severe sepsis. A prospective study. Crit Care. 2015;19(1): Johansson PI, Haase N, Perner A, Ostrowski SR. Association between sympathoadrenal activation, fibrinolysis and endothelial damage in septic patients: a prospective study. J Crit Care. 2014;29(3): Johansen ME, Johansson PI, Ostrowski SR, Bestle MH, Hein L, Jensen ALG, Soe-Jensen P, Andersen MH, Steensen M, Mohr T, et al. Profound endothelial damage predicts impending organ failure and death in sepsis. Semin Thromb Hemost. 2015;41(1): HayashidaK,ChenY,BartlettAH,ParkPW.Syndecan-1isaninvivo suppressor of Gram-positive toxic shock. J Biol Chem. 2008;283(29): Steppan J, Hofer S, Funke B, Brenner T, Henrich M, Martin E, Weitz J, Hofmann U, Weigand MA. Sepsis and major abdominal surgery lead to flaking of the endothelial glycocalix. J Surg Res. 2011;165(1): Connolly-Andersen AM, Thunberg T, Ahlm C. Endothelial activation and repair during hantavirus infection: association with disease outcome. Open Forum Infect Dis. 2014;1(1):ofu Schmidt EP, Overdier KH, Sun X, Lin L, Liu X, Yang Y, Ammons LA, Hiller TD, Suflita MA, Yu Y, et al. Urinary glycosaminoglycans predict outcomes in septic shock and ARDS. Am J Respir Crit Care Med. 2016;194: Rehm M, Bruegger D, Christ F, Conzen P, Thiel M, Jacob M, Chappell D, Stoeckelhuber M, Welsch U, Reichart B, et al. Shedding of the endothelial glycocalyx in patients undergoing major vascular surgery with global and regional ischemia. Circulation. 2007;116(17): Bruegger D, Rehm M, Abicht J, Paul JO, Stoeckelhuber M, Pfirrmann M, Reichart B, Becker BF, Christ F. Shedding of the endothelial glycocalyx during cardiac surgery: on-pump versus off-pump coronary artery bypass graft surgery. J Thorac Cardiovasc Surg. 2009;138(6): Schochl H, Cadamuro J, Seidl S, Franz A, Solomon C, Schlimp CJ, Ziegler B. Hyperfibrinolysis is common in out-of-hospital cardiac arrest: results from a prospective observational thromboelastometry study. Resuscitation. 2013; 84(4): Grundmann S, Fink K, Rabadzhieva L, Bourgeois N, Schwab T, Moser M, Bode C, Busch HJ. Perturbation of the endothelial glycocalyx in post cardiac arrest syndrome. Resuscitation. 2012;83(6): Nielsen N, Wetterslev J, Cronberg T, Erlinge D, Gasche Y, Hassager C, Horn J, Hovdenes J, Kjaergaard J, Kuiper M, et al. Targeted temperature management at 33 degrees C versus 36 degrees C after cardiac arrest. N Engl J Med. 2013;369(23): Johansson PI, Bro-Jeppesen J, Kjaergaard J, Wanscher M, Hassager C, Ostrowski SR. Sympathoadrenal activation and endothelial damage are inter correlated and predict increased mortality in patients resuscitated after out-of-hospital cardiac arrest: a post hoc sub-study of patients from the TTM-trial. PLoS One. 2015;10(3):e Jung C, Fuernau G, Muench P, Desch S, Eitel I, Schuler G, Adams V, Figulla HR, Thiele H. Impairment of the endothelial glycocalyx in cardiogenic shock and its prognostic relevance. Shock. 2015;43(5): XuL,YuWK,LinZL,TanSJ,BaiXW,DingK,LiN.Chemical sympathectomy attenuates inflammation, glycocalyx shedding and coagulation disorders in rats with acute traumatic coagulopathy. Blood Coagul Fibrinolysis. 2015;26: Opal SM, Dellinger RP, Vincent JL, Masur H, Angus DC. The next generation of sepsis clinical trial designs: what is next after the demise of recombinant human activated protein C?*. Crit Care Med. 2014;42(7): Angus DC. The search for effective therapy for sepsis: back to the drawing board? JAMA. 2011;306(23): Bernard SA, Gray TW, Buist MD, Jones BM, Silvester W, Gutteridge G, Smith K. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med. 2002;346(8): The Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med. 2002;346(8): Villablanca PA, Makkiya M, Einsenberg E, Briceno DF, Panagiota C, Menegus M, Garcia M, Sims D, Ramakrishna H. Mild therapeutic hypothermia in patients resuscitated from out-of-hospital cardiac arrest: a meta-analysis of randomized controlled trials. Ann Card Anaesth. 2016;19(1): Johansson PI, Sørensen AM, Larsen CF, Windeløv NA, Stensballe J, Perner A, Rasmussen LS, Ostrowski SR. Low hemorrhage-related mortality in trauma patients in a Level I Trauma Centre employing transfusion packages and early thrombelastography-directed hemostatic resuscitation with plasma and platelets. Transfusion. 2013;53(12): Johansson PI, Oliveri R, Ostrowski SR. Hemostatic resuscitation with plasma and platelets in trauma. A meta-analysis. J Emerg Trauma Shock. 2012;5(2): Johansson PI, Stensballe J, Oliveri R, Wade CE, Ostrowski SR, Holcomb JB. How I treat patients with massive hemorrhage. Blood. 2014;124(20): Holcomb JB, Tilley BC, Baraniuk S, Fox EE, Wade CE, Podbielski JM, del Junco DJ, Brasel KJ, Bulger EM, Callcut RA, et al. Transfusion of plasma, platelets, and red blood cells in a 1:1:1 vs a 1:1:2 ratio and mortality in patients with severe trauma: the PROPPR randomized clinical trial. JAMA. 2015;313(5): Gonzalez E, Moore EE, Moore HB, Chapman MP, Chin TL, Ghasabyan A, Wohlauer MV, Barnett CC, Bensard DD, Biffl WL, et al. Goal-directed hemostatic resuscitation of trauma-induced coagulopathy: a pragmatic randomized clinical trial comparing a viscoelastic assay to conventional coagulation assays. Ann Surg. 2015;263(6): Norris RM, Clarke ED, Sammel NL, Smith WM, Williams B. Protective effect of propranolol in threatened myocardial infarction. Lancet. 1978;2(8096): Chatterjee S, Chaudhuri D, Vedanthan R, Fuster V, Ibanez B, Bangalore S, Mukherjee D. Early intravenous beta-blockers in patients with acute coronary syndrome a meta-analysis of randomized trials. Int J Cardiol. 2013;168(2): Morelli A, Ertmer C, Westphal M, Rehberg S, Kampmeier T, Ligges S, Orecchioni A, D Egidio A, D Ippoliti F, Raffone C, et al. Effect of heart rate control with esmolol on hemodynamic and clinical outcomes in patients with septic shock: a randomized clinical trial. JAMA. 2013;310(16):

82 Clinical Science (2013) 125, (Printed in Great Britain) doi: /CS G16R single nucleotide polymorphism but not haplotypes of the β 2 -adrenergic receptor gene alters cardiac output in humans Kim Z. ROKAMP, Jonatan M. STAALSOE, Martin GARTMANN, Anna SLETGAARD, Nicolai B. NORDSBORG, Niels H. SECHER, Henning B. NIELSEN and Niels V. OLSEN Clinical Science Departments of Anaesthesia and Neuroanaesthesia, Rigshospitalet, University of Copenhagen, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark Department of Neuroscience and Pharmacology, University of Copenhagen, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark Department of Nutrition, Physical Activity and Sport Sciences, University of Copenhagen, 9 Blegdamsvej, DK-2100 Copenhagen, Denmark Abstract Variation in genes encoding the β 2 -adrenergic receptor (ADRB2) and angiotensin-converting enzyme (ACE) may influence Q (cardiac output). The 46G> A (G16R) SNP (single nucleotide polymorphism) has been associated with β 2 -mediated vasodilation, but the effect of ADRB2 haplotypes on Q has not been studied. Five SNPs within ADRB2 (46G > A, 79C > G, 491C > T, 523C > A and 1053G > C by a pairwise tagging principle) and the I/D (insertion/ deletion) polymorphism in ACE were genotyped in 143 subjects. Cardiovascular variables were evaluated by the Model flow method at rest and during incremental cycling exercise. Only the G16R polymorphism was associated with Q. In carriers of the Arg 16 allele, Q rest (resting Q ) was 0.4 [95 % CI (confidence interval), ] l/min lower than in G16G homozygotes (P = 0.048). During exercise, the increase in Q was by 4.7 (95 % CI, ) l/min per litre increase in pulmonary Vo 2 (oxygen uptake) in G16G subjects, but the increase was 0.5 ( ) l/min lower in Arg 16 carriers (P = 0.035). A similar effect size was observed for the Arg 16 haplotypes ACCCG and ACCCC. No interaction was found between ADRB2 and ACE polymorphisms. During exercise, the increase in Q was 0.5 (CI, ) l/min greater in ACE I/I carriers compared with I/D and D/D subjects (P = 0.054). In conclusion, the ADRB2 Arg 16 allele in humans is associated with a lower Q both at rest and during exercise, overriding the effects of haplotypes. Key words: angiotensin-converting enzyme (ACE), β 2 -adrenergic receptor (ADRB2), cardiac output, exercise, I/D polymorphism INTRODUCTION The β 2 -AR (β 2 -adrenergic receptor), encoded by an intronless gene (ADRB2) located on chromosome 5 (5q31-32), encompasses several SNPs (single nucleotide polymorphisms) [1 4] that appear relevant to cardiovascular regulation at rest and during exercise and implicates progression of CVD (cardiovascular disease). In the coding region of ADRB2, four SNPs are non-synonymous polymorphisms resulting in amino acid variations G16R, Q27E, V34M and T164I. The polymorphism at codon 34 is rare and with no apparent functional consequence [5]. In contrast, in vitro studies indicate that polymorphisms at codons 16 and 27 affect agonist-induced down-regulation of the receptor and that polymorphisms at codon 164 alter the coupling properties to the intracellular G α s-protein [1,2,4]. The rare Ile 164 allele is associated with rapid progression of heart failure [6]. The G16R and Q27E polymorphisms by themselves or in combination segregate with hypertension, asthma and obesity [1,2,4,7]. In addition, homozygote Gly 16 subjects demonstrate a larger Q (cardiac output) both at rest and during exercise compared with homozygote Arg 16 subjects [8], maybe because of a higher receptor density [9]. The Arg 16 allele is associated with enhanced agonist-mediated desensitization [10] and attenuated blood flow during infusion of a β-agonist in the brachial artery [11]. In contrast, the Glu 27 allele is associated with increased agonist-mediated responsiveness [10]. The role of haplotypes within ADRB2 is, however, not known, but it may be that the Gly 16 and Glu 27 haplotype promotes a more favourable cardiovascular response to exercise [3]. Abbreviations: ACE, angiotensin-converting enzyme; β 2 -AR (ADRB2), β 2 -adrenergic receptor; BMI, body mass index; BSA, body surface area; D, deletion; I, insertion; Q, cardiac output; Q rest, resting Q ; HR, heart rate; MAP, mean arterial pressure; SNP, single nucleotide polymorphism; SV, stroke volume; SVR, systemic vascular resistance; Vo 2, oxygen uptake; V o 2max, maximum V o 2. Correspondence: Dr Niels V. Olsen ( nvolsen@sund.ku.dk)

83 K.Z. Rokamp and others Polymorphism in the ACE (angiotensin-converting enzyme) gene may also affect cardiovascular regulation during exercise. At intron 16 ACE is characterized by the presence (insertion, I allele) or absence (deletion, D allele) of a 287-bp Alu repeat sequence [12]. ACE degrades vasodilating kinins and generates AngII (angiotensin II). The I/D polymorphism is a consistent marker for ACE activity in Caucasians [13]. The I/D polymorphism explains approximately one-third of the variability in plasma ACE, with circulating levels being higher in D/D carriers than in subjects with the I/I genotype [13]. The ACE I/I genotype is reported to be over-represented in elite athletes and the I allele has been linked with endurance performance [14 16], suggesting that the ACE I/D polymorphism may influence Q during exercise. Compared with a study of individual SNPs, involvement of haplotypes better reveals biological effects caused by the interaction of multiple polymorphisms [17]. In the present study, we used a pairwise tagging principle to select five marker SNPs within ADRB2 (G16R, Q27E, T164I, R175R and G351G) for construction of ADRB2 haplotypes [18]. By this approach we tested whether common human ADRB2 genotypes and haplotypes as well as the ACE I/D polymorphism by itself or by interaction with ADRB2 are associated with Q at rest and during cycling exercise. MATERIALS AND METHODS Subjects At total of 143 subjects were included in the study following written informed consent as approved by the local ethical committee (J.nr. H ) and the Danish Data Protection Agency (J.nr ). Subjects were studied after an overnight fast. Strenuous exercise was not allowed 24 h prior to the study and the subjects refrained from caffeine intake on the day of the experiment. Cardiovascular variables were measured at rest (n = 143), during submaximal exercise (n = 69) and at maximal exercise (n = 87). The subjects rested supine for 15 min after which HR (heart rate), MAP (mean arterial pressure), cardiac SV (stroke volume), and thus Q were assessed non-invasively as an average over 30 s using Modelflow methodology (Nexfin; BMEYE). Subjects hereafter exercised on a cycle ergometer (Model EC04; epower Technology ApS) and Vo 2 (oxygen uptake) was determined during submaximal and maximal exercise. Cycling exercise was performed on a custom-made etenzor bike (epower Technology ApS) securing that power was maintained irrespective of the self-chosen cadence, while ventilatory variables were determined breath-tobreath (Masterscreen CPX; VIASYS Healthcare). For evaluation at maximal exercise (n = 87), the initial workload was set at 75 W and then increased by 25 W/min until exhaustion while the subjects were encouraged to keep the pedalling rate. Subjects evaluated at submaximal exercise (n = 69) worked for 6 min at workloads of 75, 125 and W, the last workload depending on the self-reported exhaustion level during the first two work periods (intensity level 1 10). The exercise periods were separated by 5 min of recovery with the subject sitting on the ergometer. Subjects choose a cadence of rounds/min and kept that cadence for 1 min after initiating any given workload. After the submaximal exercise protocol, the subjects rested for min while sitting on the ergometer before the V o 2max (maximum V o 2 ) test. Blood samples in EDTA vacuum tubes were obtained from an arm vein for determination of genotypes. Selection of ADRB2 SNPs The ADRB2 gene was located in HapMap ( org; HapMap DataRel 22/phase II Apr 07, on NCBI B36 assembly, dbsnp b126) and the SNP genotype data were analysed in Haploview 4.0 ( Five SNPs with a minor allele frequency > 1 % [46G > A (G16R), rs ; 79C > G (Q27E), rs ; 491C > T (T164I), rs ; 523C > A (R175R), rs ; and 1053G > C (G351G), rs ] were chosen to determine the most common haplotypes in Caucasians [18 20]. Purification of DNA and genotyping DNA was purified from 200 μl of frozen whole-blood samples by the magnetic-bead-based MagneSil Blood Genomic, Max Yeld System (Promega). Genotyping was performed using Taq- Man assays with the following rs and Applied Biosystems assay ID numbers: rs , c _20; rs , c _20; rs , c _20; rs , c _10; and rs , c_ _10. The assays were analysed using PCR by a fast real-time PCR device (Applied Biosystem 7500) according to the manufacturer s instructions. Genotyping of the ACE I/D polymorphism was performed in duplicate using the primers 5 -CACACCCTGAAGTACGGCAC- 3 (sense) and 5 -GTGGCCATCACATTCGTCAG-3 (antisense). As the D band is preferentially amplified with the initial primer set, resulting in some I/D genotypes being interpreted as D/D, we reanalysed the D/D genotypes with an I-specific primer [21,22]. Statistical analysis Statistical analysis was with R version (MAC GUI 1.42 Leopard build 64-bit, 5910; ANOVA was used for comparison of age, height, mass, BSA (body surface area) and BMI (body mass index) between genotypes, whereas Fisher s Exact test was used for gender. Haplotypes were estimated from genotypes using a Bayesian approach implemented with PHASE version 2.1; the programme estimates the haplotype frequencies in the sample population, the associated CI (confidence interval), and the most likely haplotype pair for each individual [23]. Genotype frequencies and Hardy Weinberg equilibrium were calculated with software from the R-project ( We initially used a haplotype score procedure ( haplo.stat ) to estimate the relative strength of dominant, additive or recessive models on Q rest (resting Q)and Q in relation to V o 2 during exercise ( Q/ V o 2 ). Then, genotype effects on Q rest were tested with linear models, whereas haplotype effects were evaluated by the use of the haplo.glm procedure of the Haplo Stat package that allows for testing a genotype matrix with generalized linear models of dominant, additive or recessive haplotype effects [23]. 192 C The Authors Journal compilation C 2013 Biochemical Society

84 Cardiac output and variation in ADRB2 and ACE Table 1 Frequencies of SNPs in the ADRB2 gene and the I/D polymorphism in the ACE gene in 143 Danish Caucasians Genome frequencies are for major allele homozygotes/heterozygotes/minor allele homozygotes. SNP position Major/minor base Amino acid coding and position Genotype frequencies Minor allele frequency rs G > A G16R 0.32/0.52/ rs C > G Q27E 0.29/0.57/ rs C > T T164I 0.97/0.03/ rs C > A R175R 0.73/0.26/ rs G > C G351G 0.58/0.34/ I/D 0.26/0.44/ Table 2 Arrangement of SNPs in haplotypes within the ADRB2 gene in 143 Danish Caucasians Haplotype frequencies (means + S.D.) were derived from PHASE 2.1. Nucleotide position... 46G > A 79C > G 491C > T 523C > A 1053G > C rs identification of SNP... rs rs rs rs rs Haplotype number Amino acid location... G16R Q27E T164I R175R G351G Frequency (%) 1 G G C C G A C C C G G C C A C A C C C C G C C C C G C T A C G G C C C G C C C G A C C A G The Q/ V o 2 relationship is Q = Q rest + Q/ V o 2 V o 2 + a random error. Thus Q rest is the intercept, Q/ V o 2 is the slope of the line, and Q increases linearly with V o 2 during submaximum exercise [24]. The Q/ V o 2 relationship was tested under a mixed effects regression model ( nlme ) with effects of the ADRB2 Arg 16 allele, BSA and age on both the intercept and slope (fixed effects). Subjects were included in the model as a random effect, i.e. that each subject has his/her own regression line with individual intercept and slope. Maximum likelihood generalized least-squares estimates were used throughout. Model assumptions were checked by examining residual plots, which in all cases were assessed to be close to normality. The method of covariate selection was to first include all biologically relevant confounders, i.e. sex, age and BSA, with two-way interactions and secondly proceed to the stepwise removal of effects with non-significant impact on the model fit. Cardiovascular and genetic data were analysed independently. RESULTS A total of 143 (59 women) healthy non-smoking Caucasians entered the study (age, years; height, cm; weight, kg; BSA, m 2 ; BMI, kg/m 2 ;values are means + S.D.). Frequencies of ADRB2 and ACE genotypes are shown in Table 1. Analysis of the five marker SNPs in ADRB2 revealed the existence of nine distinct haplotypes (Table 2). The four most common haplotypes were observed in 94 % of the subjects, whereas the others each had frequencies below 3 % (Table 2). Genotype and haplotype distributions within ADRB2 and ACE were in close agreement with those observed in Caucasians [13,18 20,25,26]. All genotypes and haplotypes were within Hardy Weinberg equilibrium. There were no significant differences in sex, age, height, mass, BSA or BMI between genotypes, haplotypes or haplotype pairs. Haplotype scores for dominant, additive and recessive effects indicated a dominant effect on both Q rest (P = 0.07) and Q/ V o 2 (P = 0.047) that was associated with the ACCCG haplotype. All other haplotypes and models had lower scores that were not statistically significant. We therefore proceeded to test only dominant models. Q at rest and ADRB2 polymorphisms Cardiovascular variables at rest and during incremental cycling are shown in Figure 1. Only the G16R polymorphism was associated with Q at rest (Figure 2). Compared with R16R subjects, Q rest in G16G subjects was greater by 0.4 (CI, ) l/min (P = adjusted for additive effects of age and BSA). Without adjustment for covariates, the effect was similar (difference = 0.4 l/min;, P = 0.029, Student s t test). None of the other SNPs were associated with Q or any other cardiovascular variables. Cardiovascular variables at rest in relation to ADRB2 G16R genotypes are shown in Table 3. Highest values of HR and SV were observed in G16G subjects (Table 3), but none of the differences reached statistical significance. Influence of the four major haplotypes on Q rest is shown in Table 4. The haplotypes ACCCG (#2) and ACCCC (#4) encompassing the Arg 16 allele had similar effect sizes as when the G16R polymorphism was 193

85 K.Z. Rokamp and others Figure 2 Mean values of Q rest (A) and the relationship between changes in Q and V o 2 ( Q/ V o 2 ) (B) according to genotypes of the ADRB2 G16R polymorphism and the ACE I/D polymorphism Error bar denotes 95 % CIs. P = (Arg 16 dominant model, age and mass-adjusted ANOVA). #P = (Arg 16 dominant model, age and mass-adjusted mixed effects model). Figure 1 Q, V o 2, HR, SV, MAP and SVR at rest and during incremental cycling exercise at 75 W, 125, W and at maximal workload Solid line denotes the median, the box the interquartile range and whiskers extend to the most extreme data point which is no more than the 1.5 interquartile range from the box. tested alone, although not statistically significant when adjusted for age and BSA. Q during incremental cycling and ADRB2 polymorphisms The increase in Q during incremental exercise (n = 69) depended on age, BSA and the G16R polymorphism (Table 5 and Figure 2). Carriers of G16G increased their Q by 4.7 l/min (CI, ) for every 1 l/min V o 2 increase, whereas Arg 16 carriers of the same ageandbsahada Q of 4.2 l/min (CI, ) (Table 5 and Figure 2). In this model, the Arg 16 allele did not by itself exert a statistically significant effect on the intercept ( Q rest ). During maximal exercise there were no difference in V o 2max or cardiovascular variables between genotypes. Throughout incremental exercise, highest values of MAP, HR and SV and lowest values of SVR (systemic vascular resistance) were observed in G16G subjects (results not shown), but none of the differences reached statistical significance. The haplotype dominant model (Table 4) revealed a borderline significant effect of ACCCG (#2) on Q/ V o 2 compared with the reference haplotype GGCCG (#1) (P = 0.056), with an estimated difference in Q/ V o 2 of 0.5 (CI, ) l/min. This effect is of the same magnitude as observed for the Arg 16 dominant genotype. As a difference in effects on Q/ V o 2 between ACCCG and ACCCC may reflect a haplotype-specific effect separate from the G16R effect, we proceeded to analyse a subset of 194 C The Authors Journal compilation C 2013 Biochemical Society

86 Cardiac output and variation in ADRB2 and ACE Table 3 Cardiovascular variables at rest in relation to the G16R polymorpism in the ADRB2 gene Values are means + S.D. P = compared with R16R adjusted for additive effects of age and BSA. G16R genotype (n = 143) G16G G16R R16R Parameter (n = 46) (n = 75) (n = 22) Systolic pressure (mmhg) Diastolic pressure (mmhg) MAP (mmhg) HR (beats/min) SV (ml) Q (l/min) SVR (dyn s cm 5 ) Vo 2 (l/kg of body mass) Table 6 Cardiovascular variables at rest in relation to genotypes of the I/D polymorphism of the ACE gene Values are means + S.D. Genotype (n = 143) I/I I/D D/D Parameter (n = 37) (n = 63) (n = 43) Systolic pressure (mmhg) Diastolic pressure (mmhg) MAP (mmhg) HR (beats/min) SV (ml) Q (l/min) SVR (dyn s cm 5 ) Vo 2 (l/kg of body mass) Table 4 ADRB2 gene haplotypes, Q rest and the relationship between changes in Q and V o 2 ( Q/ V o 2 ) Estimated effects of a given haplotype on Q rest and Q / V o 2 during exercise with 95 % CI in parentheses for a 30-year-old subject with BSA of 1.9 m 2. n = 143 for Q rest ; n = 69 for Q / V o 2. P value (of t-statistic) of the difference from the haplotype GGCGG in a generalized linear model of haplotypes corrected for age and BSA. Haplotype Q rest P value Q/ V o 2 P value GGCCG 6.4 ( ) 4.8 ( ) GCCAC 6.4 ( ) ( ) 0.26 ACCCG 6.0 ( ) ( ) ACCCC 6.1 ( ) ( ) Others 6.5 ( ) ( ) All other haplotypes are accounting for a total of 5 % of the haplotypes. Table 5 Predictive effect of the ADRB2 G16R polymorphism on Q and relationship between changes in Q and V o 2 ( Q/ V o 2 ) The mean (95 % CI) effect is given. Population effects of a linear mixed model by which an exercising individual of 40 years of age, BSA = 2.0 m 2 and carrier of at least one Arg 16 allele who increases his V o 2 by 1 l/min can be estimated to increase his Q by Q / V o 2 = ( 0.49) + (1 0.48) + (1 0.19) = 3.96 l/min. n = 69. Model Effect (l/min) P value Q rest (model intercept) 6.4 (5.9 to 6.9) Q/ V o 2 (model slope) 4.7 (4.3 to 5.2) Q rest (intercept) effects Arg 16 allele 0.1 ( 0.6 to 0.3) Age (per decade after 30 years) 0.5 ( 0.7 to 0.2) BSA (per 0.1 change from 1.9 m 2 ) 0.9 ( 0.2 to 2.0) Q / V o 2 (slope) effects Arg 16 allele 0.5 ( 0.9- to 0.0) Age (per decade after 30 years) 0.5 ( 0.7- to 0.2) BSA (per 0.1 change from 1.9 m 2 ) 0.2 ( 0.3- to 0.1) Q rest for an individual with age = 30 years, BSA = 1.9 m 2 and resting V o 2 = l/min. Effects on intercept are additive. Increase in Q in l/min per increase in Vo 2 of 1 l/min. Effects on slope are additive. P value (of t-statistic) of zero effect size. Table 7 Predictive effect of the I/D polymorphism in the ACE gene on Q and the relationship between changes in Q and Vo 2 ( Q/ Vo 2 ) The mean (95 % CI) effect is given. n = 69. Model Effect (l/min) P value Q rest (model intercept) 6.4 (6.0 to 6.8) Q / V o 2 (model slope) 4.3 (3.9 to 4.7) Q rest (intercept) effects I/I genotype 0.3 ( 0.8 to 0.2) Age (per decade after 30) 0.5 ( 0.8 to 0.3) BSA (per 0.1 change from 1.9 m 2 ) 0.1 ( 0.0 to 0.2) Q / V o 2 (slope) effects I/I genotype 0.5 ( ) Age (per decade after 30 years) 0.5 ( 0.7 to 0.2) BSA (per 0.1 change from 1.9 m 2 ) 0.2 ( 0.3 to 0.1) Q rest for an individual with D/D or I/D genotype, age = 30 years, BSA = 1.9 m 2 and V o 2 = 284 l/min. Effects on intercept are additive. Increase of Q in l/min per increase in V o 2 of 1 l/min. Effects on slope are additive. P value (of t-statistic) of zero effect size. the data for subjects expressing either ACCCG or ACCCC and tested these differences against each other. ACCCG haplotypes (n = 32) had a Q/ V o 2 of 0.3 (CI, ) l/min lower than the ACCCC haplotypes (n = 12) (P = 0.33). Thus, the data do not support the presence of a haplotype-specific effect on Q/ V o 2. ACE I/D polymorphism and Q Q rest in I/I genotypes did not differ from the values in I/D and D/D carriers (P = 0.64) (Table 6 and Figure 2). During incremental exercise Q/ V o 2 was 0.5 ( ) l/min higher in I/I genotypes compared with I/D and D/D genotypes (Table 7 and Figure 2). Without adjustment for age and BSA this effect was significant (P = 0.025) and amounted to 0.7 ( 0.1 to 1.1) l/min. Throughout incremental exercise, the highest values of MAP, HR and SV and lowest values of SVR were observed in I/I subjects (results not shown), but the differences did not reach statistical significance. During maximal exercise there were no difference in V o 2max or cardiovascular variables between ACE I/D genotypes. We evaluated both Q rest and Q/ V o 2 during exercise 195

87 K.Z. Rokamp and others with the inclusion of interactive effects between ACE I/D and ADRB2 G16R; however, none were significant (P = 0.5 for interaction on Q/ V o 2 and P = 0.21 for interaction on Q rest ). In subjects with the ADRB2 G16G genotype, the presence of the ACE I/I genotype (n = 14) did not increase Q compared with the presence of the D/D genotype (n = 11; P = 0.26). DISCUSSION By the use of five marker SNPs in ADRB2, the present study in Scandinavian Caucasians identified four major haplotypes observed in 95 % of the subjects. The data indicate that only the G16R SNP in ADRB2 is associated with Q, independent of arrangements in haplotypes. The G16R SNP by itself produces higher values of Q rest and Q/ V o 2 in G16G homozygotes compared with G16R heterozygotes and R16R homozygotes. In addition, the study suggests that carriers of the ACE I/I genotype may have higher Q/ V o 2 during incremental exercise than I/D and D/D subjects. The presence of the ACE I/I genotype, however, did not enhance the effect of the ADRB2 G16G genotype. The results of the present study are in line with studies that have evaluated the effect of the G16R SNP on the cardiovascular response to exercise. In response to isometric handgrip, individuals homozygous for Gly 16 showed a greater increase in HR and Q compared with subjects homozygous for Arg 16 [27]. Homozygote Gly 16 subjects demonstrated an increased cardiac SV and Q both at rest and during exercise compared with homozygote Arg 16 subjects [8]. In addition, Gly 16 homozygotes had a greater fractional shortening, ejection fraction and midwall shortening compared with both heterozygotes and Arg 16 homozygotes [28]. In the present study, the G16G homozygotes had a higher Q rest compared with G16R heterozygotes and R16R homozygotes, but the difference in Q was smaller than reported by Snyder et al. [8]. The group of subjects in the two studies differs in that our group had a lower BMI and higher V o 2max. Furthermore, the age range was wider in our group of participants. There were also differences in methods of measuring Q. Snyder et al. [8] used the open-circuit acetylene uptake method, whereas we used the Modelflow method [29]. The Modelflow method seems to underestimate the increase in Q during heat stress [30], but it has been successfully validated against a thermodilution estimate during a deliberate reduction in central blood volume induced by standing up in healthy subjects [31], during cardiac surgery [32], in intensive care medicine [33] and during liver transplantation surgery [34]. Other SNPs in ADRB2 possess a potential effect on cardiovascular regulation. The Q27E SNP has been linked with increased agonist-mediated responsiveness [10]. Subjects homozygous for Gly 16 and Glu 27 demonstrated a greater forearm vasodilator response to mental stress and isometric handgrip, an effect that was attributed to position 27, since Gly 16 + Glu 27 subjects had a greater response than Gly 16 + Gln 27 and Arg 16 + Gln 27 subjects [35]. Snyder et al. [8] found that subjects homozygote for Gly 16 had a larger Q than Arg 16 homozygotes, but among G16G subjects Q differed significantly according to the Q27E genotype, with Glu 27 homozygotes having the lowest Q. According to Snyder et al. [3] the primary difference in phenotype seems to be weighted by amino acids 16 and 27 (G16R and Q27E). In another study, the Gly 16 and Glu 27 haplotype combination has been suggested as a marker for talent identification of athletes [36]. We included haplotypes in our analysis, but did not find any significant effect of the combination of G16R and Q27E. Rather our data suggest that the G16R SNP alone seems to be responsible for the effect on Q both at rest and during exercise. The ACE I/I genotype has been observed to be overrepresented in elite athletes and a positive association of the I allele with elite endurance performance has been suggested [16]. The genotypic effects on physiological phenotypes may be mediated by variation in the levels of ACE activity associated with the ACE I/D polymorphism [37]. Montgomery et al. [15] found LV (left ventricular) mass associated with the ACE D allele. In addition, the ACE D/D genotype has been linked with a small V o 2max [38,39]. However, other studies found no such relationship [40,41]. The results of the present study suggest that the ACE I/D polymorphism has no effect on Q rest, but during exercise the I/I genotype subjects had increased Q/ V o 2 compared with I/D and D/D subjects. Although this may support a beneficial effect of the I allele in promoting athletic performance, there was no difference in V o 2max between the ACE I/D genotypes. Moreover, we show that the presence of the ACE I/I genotype does not enhance the effect of the ADRB2 Gly/Gly genotype. The present study has several limitations. First, the sample size was small, especially when the population was divided into subgroups, and the level of significance appeared to be borderline. The present study did not have enough statistical power to distinguish between the effect of the ADRB2 G16R polymorphism on HR and SV, and only the overall effect on Q was recognized. Secondly, circulating levels of catecholamines were not measured and it is possible that differences in receptor agonist concentrations influenced the results. The exact molecular mechanism of variation in ADRB2 remains unknown but may include unique interaction between genotypes in conformation and downstream signalling of β 2 -AR [2,4]. In conclusion, the results of the present study confirm that the ADRB2 G16R polymorphism is associated with a higher Q rest in G16G homozygotes compared with G16R heterozygotes and R16R homozygotes. Furthermore, during light-to-moderate exercise, G16G carriers demonstrate an enhanced increase in Q for a given increase in V o 2 compared with Arg 16 homozygotes. Analysis of ADRB2 haplotypes indicates that only the G16R SNP has such an effect, overriding the impact of haplotypes. The ACE I/I genotype may also augment the exercise-induced increase in Q but does not interact with the effect of the ADRB2 G16R polymorphism. CLINICAL PERSPECTIVES Endurance performance is highly dependent upon adequate increases in Q and oxygen delivery. In part, inter-individual differences in cardiovascular responses to exercise may be caused by genetic variation in ADRB C The Authors Journal compilation C 2013 Biochemical Society

88 Cardiac output and variation in ADRB2 and ACE We found that the G16R polymorphism in ADRB2 by itself, independent of arrangements in haplotypes, contributes to heterogeneity in Q at rest and during incremental exercise, along with age and body surface area. The results add to the understanding of gene exercise interactions and genetic influence on sports performance. AUTHOR CONTRIBUTION Kim Zillo Rokamp participated in study design, collected the data, performed data analysis and wrote the first draft of the paper. Jonatan Staalsoe performed data analysis and contributed to preparation of the paper. Martin Gartmann and Anna Sletgaard participated in study design, collected the data and contributed to preparation of the paper. Nicolai Nordsborg, Niels Secher, Henning Nielsen and Niels Olsen participated in study design, performed data analysis and contributed to preparation of the paper. FUNDING The study was supported by the Scientific Board of the Rigshospitalet, Copenhagen, Denmark, by the Danish Society of Anaesthesiology and by the Ehrenreich Foundation, Denmark. REFERENCES 1 Leineweber, K., Büscher, R., Bruck, H. and Brodde, O.-E. (2004) β-adrenoceptor polymorphisms. Naunyn-Schmiedebergs Arch. Pharmacol. 369, Small, K. M., McGraw, D. W. and Liggett, S. B. (2003) Pharmacology and physiology of human adrenergic receptor polymorphisms. Annu. Rev. Pharmacol. Toxicol. 43, Snyder, E. M., Johnson, B. D. and Joyner, M. J. (2008) Genetics of β 2 -adrenergic receptors and the cardiopulmonary response to exercise. Exercise Sport Sci. Rev. 36, Zaugg, M. and Schaub, M. C. (2005) Genetic modulation of adrenergic activity in the heart and vasculature: implications for perioperative medicine. Anesthesiology 102, Green, S. A., Turki, J., Hall, I. P. and Liggett, S. B. (1995) Implications of genetic variability of human β 2 -adrenergic receptor structure. Pulm. Pharmacol. 8, Liggett, S. B., Wagoner, L. E., Craft, L. L., Hornung, R. W., Hoit, B. D., McIntosh, T. C. and Walsh, R. (1998) The Ile 164 β 2 -adrenergic receptor polymorphism adversely affects the outcome of congestive heart failure. J. Clin. Invest. 102, Beuther, D. A., Weiss, S. T. and Sutherland, E. R. (2006) Obesity and asthma. Am. J. Respir. Crit. Care Med. 174, Snyder, E. M., Beck, K. C., Dietz, N. M., Eisenach, J. H., Joyner, M. J., Turner, S. T. and Johnson, B. D. (2006) Arg16Gly polymorphism of the β 2 -adrenergic receptor is associated with differences in cardiovascular function at rest and during exercise in humans. J. Physiol. 571, Snyder, E. M., Hulsebus, M. L., Turner, S. T., Joyner, M. J. and Johnson, B. D. (2006) Genotype related differences in β 2 adrenergic receptor density and cardiac function. Med. Sci. Sports Exercise 38, Dishy, V., Sofowora, G. G., Xie, H. G., Kim, R. B., Byrne, D. W., Stein, C. M. and Wood, A. J. (2001) The effect of common polymorphisms of the β 2 -adrenergic receptor on agonistmediated vascular desensitization. N. Engl. J. Med 345, Garovic, V. D., Joyner, M. J., Dietz, N. M., Boerwinkle, E. and Turner, S. T. (2002) β 2 -Adrenergic receptor polymorphism and nitric oxide-dependent forearm blood flow responses to isoproterenol in humans. J. Physiol. 546, Rigat, B., Hubert, C., Alhenc-Gelas, F., Cambien, F., Corvol, P. and Soubrier, F. (1990) An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J. Clin. Invest. 86, Agerholm-Larsen, B., Tybjaerg-Hansen, A., Schnohr, P. and Nordestgaard, B. G. (1999) ACE gene polymorphism explains % of variability in serum ACE activity in both women and men in the population at large: the Copenhagen City Heart Study. Atherosclerosis 147, Gayagay, G., Yu, B., Hambly, B., Boston, T., Hahn, A., Celermajer, D. S. and Trent, R. J. (1998) Elite endurance athletes and the ACE I allele the role of genes in athletic performance. Hum. Genet. 103, Montgomery, H. E., Clarkson, P., Dollery, C. M., Prasad, K., Losi, M. A., Hemingway, H., Statters, D., Jubb, M., Girvain, M., Varnava, A. et al. (1997) Association of angiotensin-converting enzyme gene I/D polymorphism with change in left ventricular mass in response to physical training. Circulation 96, Myerson, S., Hemingway, H., Budget, R., Martin, J., Humphries, S. and Montgomery, H. (1999) Human angiotensin I-converting enzyme gene and endurance performance. J. Appl. Physiol. 87, Little, J., Higgins, J. P. T., Ioannidis, J. P. A., Moher, D., Gagnon, F., von Elm, E., Khoury, M. J., Cohen, B., Davey-Smith, G., Grimshaw, J. et al. (2009) Strengthening the reporting of genetic association studies (STREGA): an extension of the strengthening the reporting of observational studies in epidemiology (STROBE) statement. J. Clin. Epidemiol. 62, Jensen, M. K., Nielsen, M., Koefoed, P., Nielsen, H. B., Ullum, H., Haastrup, E., Romner, B., Moltke, F. B. and Olsen, N. V. (2009) Haplotype structure of the β 2 -adrenergic receptor gene in 814 Danish Caucasian subjects and association with body mass index. Scand. J. Clin. Lab. Invest. 69, Drysdale, C. M., McGraw, D. W., Stack, C. B., Stephens, J. C., Judson, R. S., Nandabalan, K., Arnold, K., Ruano, G. and Liggett, S. B. (2000) Complex promoter and coding region β 2 -adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness. Proc. Nat. Acad. Sci. U.S.A. 97, Gjesing, A. P., Sparsø, T., Borch-Johnsen, K., Jørgensen, T., Pedersen, O., Hansen, T. and Olsen, N. V. (2009) No consistent effect of ADRB2 haplotypes on obesity, hypertension and quantitative traits of body fatness and blood pressure among 6,514 adult Danes. PLoS ONE 4, e Shanmugam, V., Sell, K. W. and Saha, B. K. (1993) Mistyping ACE heterozygotes. PCR Methods Appl. 3, Suehiro, T., Morita, T., Inoue, M., Kumon, Y., Ikeda, Y. and Hashimoto, K. (2004) Increased amount of the angiotensinconverting enzyme (ACE) mrna originating from the ACE allele with deletion. Hum. Genet. 115, Stephens, M., Smith, N. J. and Donnelly, P. (2001) A new statistical method for haplotype reconstruction from population data. Am. J. Hum. Genet. 68, Yamaguchi, I., Komatsu, E. and Miyazawa, K. (1986) Intersubject variability in cardiac output-o 2 uptake relation of men during exercise. J. Appl. Physiol. 61, Large, V., Hellström, L., Reynisdottir, S., Lönnqvist, F., Eriksson, P., Lannfelt, L. and Arner, P. (1997) Human β 2 adrenoceptor gene polymorphisms are highly frequent in obesity and associate with altered adipocyte β 2 adrenoceptor function. J. Clin. Invest. 100,

89 K.Z. Rokamp and others 26 Staalsø, J. M., Nielsen, M., Edsen, T., Koefoed, P., Springborg, J. B., Moltke, F. B., Laursen, H., Nielsen, H. B. and Olsen, N. V. (2011) Common variants of the ACE gene and aneurysmal subarachnoid hemorrhage in a Danish population: a case-control study. J. Neurosurg. Anesthesiol. 23, Eisenach, J. H., Barnes, S. A., Pike, T. L., Sokolnicki, L. A., Masuki, S., Dietz, N. M., Rehfeldt, K. H., Turner, S. T. and Joyner, M. J. (2005) Arg 16 /Gly β 2 -adrenergic receptor polymorphism alters the cardiac output response to isometric exercise. J. Appl. Physiol. 99, Tang, W. (2003) The Arg16Gly polymorphism of the β 2 -adrenergic receptor and left ventricular systolic function. Am. J. Hypertens. 16, Bogert, L. W. J. and van Lieshout, J. J. (2005) Non-invasive pulsatile arterial pressure and stroke volume changes from the human finger. Exp. Physiol. 90, Shibasaki, M., Wilson, T. E., Bundgaard-Nielsen, M., Seifert, T., Secher, N. H. and Crandall, C. G. (2011) Modelflow underestimates cardiac output in heat-stressed individuals. Am. J. Physiol. Reg. Integr. Comp. Physiol. 300, R486 R Harms, M. P., Wesseling, K. H, Pott, F., Jenstrup, M., Van Goudoever, J., Secher, N. H. and Van Lieshout, J. J. (1999) Continuous stroke volume monitoring by modelling flow from non-invasive measurement of arterial pressure in humans under orthostatic stress. Clin. Sci. 97, Jansen, J. R., Schreuder, J. J., Mulier, J. P., Smith, N. T., Settels, J. J. and Wesseling, K. H. (2001) A comparison of cardiac output derived from the arterial pressure wave against thermodilution in cardiac surgery patients. Br. J. Anaesth. 87, Jellema, W. T., Wesseling, K. H., Groeneveld, A. B., Stoutenbeek, C. P., Thijs, L. G. and van Lieshout, J. J. (1999) Continuous cardiac output in septic shock by simulating a model of the aortic input impedance: a comparison with bolus injection thermodilution. Anesthesiology 90, Nissen, P., Van Lieshout, J. J., Novovic, S., Bundgaard-Nielsen, M. and Secher, N. H. (2009) Techniques of cardiac output measurement during liver transplantation: arterial pulse wave versus thermodilution. Liver Transplant. 15, Trombetta, I. C., Batalha, L. T., Rondon, M. U. P. B., Laterza, M. C., Frazzatto, E., Alves, M. J. N. N., Santos, A. C., Brum, P. C., Barretto,A.C.P., Halpern, A. et al. (2005) Gly16 + Glu27 β 2 -adrenoceptor polymorphisms cause increased forearm blood flow responses to mental stress and handgrip in humans. J. Appl. Physiol. 98, Sarpeshkar, V. and Bentley, D. J. (2010) Adrenergic- β 2 receptor polymorphism and athletic performance. J. Hum. Genet. 55, Moran, C. N., Vassilopoulos, C., Tsiokanos, A., Jamurtas, A. Z., Bailey, M. E. S., Montgomery, H. E., Wilson, R. H. and Pitsiladis, Y. P. (2006) The associations of ACE polymorphisms with physical, physiological and skill parameters in adolescents. Eur. J. Hum. Genet. 14, Abraham, M. R. (2002) Angiotensin-converting enzyme genotype modulates pulmonary function and exercise capacity in treated patients with congestive stable heart failure. Circulation 106, Hagberg, J. M., Ferrell, R. E., McCole, S. D., Wilund, K. R. and Moore, G. E. (1998) V o 2max is associated with ACE genotype in postmenopausal women. J. Appl. Physiol. 85, Day, S. H., Gohlke, P., Dhamrait, S. S. and Williams, A. G. (2007) No correlation between circulating ACE activity and VO 2 max or mechanical efficiency in women. Eur. J. Appl. Physiol. 99, Roltsch, M. H., Brown, M. D., Hand, B. D., Kostek, M. C., Phares, D. A., Huberty, A., Douglass, L. W., Ferrell, R. E. and Hagberg, J. M. (2005) No association between ACE I/D polymorphism and cardiovascular hemodynamics during exercise in young women. Int. J. Sports Med. 26, Received 12 October 2012/4 January 2013; accepted 25 February 2013 Published as Immediate Publication 25 February 2013, doi: /CS C The Authors Journal compilation C 2013 Biochemical Society

90

91

92

93

94 Am J Physiol Regul Integr Comp Physiol 297: R1058 R1065, First published August 5, 2009; doi: /ajpregu Jugular venous pooling during lowering of the head affects blood pressure of the anesthetized giraffe E. Brøndum, 1 J. M. Hasenkam, 2 N. H. Secher, 6 M. F. Bertelsen, 7,8 C. Grøndahl, 8 K. K. Petersen, 4 R. Buhl, 7 C. Aalkjær, 1 U. Baandrup, 5,9 H. Nygaard, 2,10 M. Smerup, 2 F. Stegmann, 11 E. Sloth, 2 K. H. Østergaard, 9 P. Nissen, 6 M. Runge, 6 K. Pitsillides, 12 and T. Wang 3 1 Institute of Physiology and Biophysics, 2 Institute of Clinical Medicine, Departments of CardioThoracic and Vascular Surgery, and 3 Zoophysiology, Department of Biological Sciences, Aarhus University, Aarhus, Denmark; 4 Department of Radiology and 5 Institute of Pathology, Aarhus University Hospital, Aarhus, Denmark; 6 Department of Anesthesiology, Rigshospitalet, 7 Department of Large Animal Science, University of Copenhagen, and 8 Center for Zoo and Wild Animal Health, Copenhagen Zoo, Copenhagen, Denmark; 9 Vendsyssel Hospital, Hjørring, Denmark; 10 Engineering College of Aarhus, Aarhus, Denmark; 11 Department of Companion Animal Clinical Studies, University of Pretoria, Pretoria, South Africa; and 12 EndoSomatic Technologies LLC, Sacramento, California Submitted 26 September 2008; accepted in final form 14 July 2009 Brøndum E, Hasenkam JM, Secher NH, Bertelsen MF, Grøndahl C, Petersen KK, Buhl R, Aalkjær C, Baandrup U, Nygaard H, Smerup M, Stegmann F, Sloth E, Østergaard KH, Nissen P, Runge M, Pitsillides K, Wang T. Jugular venous pooling during lowering of the head affects blood pressure of the anesthetized giraffe. Am J Physiol Regul Integr Comp Physiol 297: R1058 R1065, First published August 5, 2009; doi: /ajpregu How blood flow and pressure to the giraffe s brain are regulated when drinking remains debated. We measured simultaneous blood flow, pressure, and cross-sectional area in the carotid artery and jugular vein of five anesthetized and spontaneously breathing giraffes. The giraffes were suspended in the upright position so that we could lower the head. In the upright position, mean arterial pressure (MAP) was mmhg (mean SE), carotid flow was l/min, and carotid cross-sectional area was cm 2. Central venous pressure (CVP) was 4 2 mmhg, jugular flow was l/min, and jugular cross-sectional area was cm 2 (n 4). Carotid arterial and jugular venous pressures at head level were and 7 4 mmhg, respectively. When the head was lowered, MAP decreased to mmhg, while carotid cross-sectional area and flow remained unchanged. Cardiac output was reduced by 30%, CVP decreased to 1 2 mmhg (P 0.01), and jugular flow ceased as the jugular cross-sectional area increased to cm 2 (P 0.01), corresponding to accumulation of 1.2 l of blood in the veins. When the head was raised, the jugular veins collapsed and blood was returned to the central circulation, and CVP and cardiac output were restored. The results demonstrate that in the upright-positioned, anesthetized giraffe cerebral blood flow is governed by arterial pressure without support of a siphon mechanism and that when the head is lowered, blood accumulates in the vein, affecting MAP. Giraffa camelopardalis, venous pressure; heart rate; flow; Starling mechanism; siphon; waterfall IT HAS LONG BEEN DEBATED how cerebral blood flow and pressure are regulated when the giraffe, within seconds, changes position from standing, with the brain elevated some 2 to 3 m above the heart, to the drinking position, where the brain is lowered some 2 m below the heart. Although experiments have shown that the adult giraffe is endowed with a mean arterial pressure (MAP) of twice the value of humans (14, 27, 36), the Address for reprint requests and other correspondence: T. Wang, Zoophysiology, Dept. of Biological Sciences, Aarhus Univ., C. F. Møllers Allé, build. 1131, 8000 Aarhus C, Denmark. ( tobias.wang@biology.au.dk). regulatory mechanisms during postural changes are not fully understood. Despite a high MAP, it has been suggested that cerebral perfusion of the upright giraffe needs to be supported by a siphon mechanism (2), where the energy needed to overcome gravity in the carotid artery is recovered as blood returns to the heart through the jugular vein. However, a siphon mechanism in the giraffes 2 m long neck would require that the jugular venous pressure at the cranial end is negative by 100 mmhg or a high pressure at the base of the jugular vein. The siphon theory has been addressed in numerous theoretical and mechanical models of which some have reached conflicting conclusions (4, 6, 19, 25, 27, 30, 33). Alternatively, the jugular veins may be collapsed, as in humans when the head is upright (10). In that case, a siphon mechanism cannot operate, and the waterfall analogy has been used to describe venous return (23). The in vivo jugular venous pressure of the giraffe has been reported to be positive close to the head and to approach zero toward the heart (18). This positive pressure gradient negates a functional siphon, but is also not consistent with the common understanding of the waterfall hypothesis (30). The largest gravitational challenge to the giraffe s circulatory system is when it lowers the head to drink and raises it again. However interesting this question is, there is only one report of distal carotid arterial pressure from one fully recovered giraffe that voluntarily lowered its head to drink (35). This report shows one brief increase to almost 350 mmhg in distal carotid pressure (35). A threefold rise in the inflow pressure to the cerebral circulation would probably overwhelm cerebral autoregulation (24, 29), as it takes about 3 s for (human) autoregulation to respond to sudden increase in MAP (31). Other mechanisms are, therefore, likely to be involved in protecting the giraffes from cerebral edema or hemorrhage. The rete mirabile, a tortuous network of small veins and arterioles derived from the carotid artery upon entering the skull (15), has been speculated to reduce pressure within the giraffe s cerebral arterial vessels by expansion of the venous vessels when the head is lowered (28). An expansion of the highly compliant jugular vein when the head is lowered has also been suggested (15), and some correlation between head posture and MAP have been shown (35). However, the changes in dimension of the jugular vein Downloaded from by on May 2, 2017 R /09 $8.00 Copyright 2009 the American Physiological Society

95 during postural changes have not been visualized, and its relation to central venous pressure (CVP) and blood pressure has not been documented. To study this, we determined changes in central and regional hemodynamic variables and made ultrasound images of the carotid artery and jugular veins in five spontaneously breathing anesthetized giraffes. The giraffes were suspended in a sling system designed to avoid pressure to the thoracic and abdominal regions as pressure in these regions was demonstrated to decrease cardiac output and carotid flow. The setup allowed us to lower and raise the head from its normal upright position to heart level by swinging the neck to one side or the other. Heart rate and blood pressure of the anesthetized giraffes were verified by telemetric measurements in one freely walking giraffe. VENOUS POOLING IN THE GIRAFFE NECK R1059 METHODS Six young male giraffes (Giraffa camelopardalis) bred for trophy hunting in Namibia were caught and transported to a quarantine facility in Hammanskraal, South Africa. The giraffes were adapted to their new setting for 20 days prior to the experiments. The experimental protocol was approved by the Danish Inspectorate for Animal Experimentation, the Danish Ministry of Justice; by the Animal Ethics Screening Committee at the University of Witwatersrand, Johannesburg; and by the Animal Use and Care Committee, University of Pretoria, South Africa. Permission to euthanize the animals after experimentation was granted by the Gauteng Province of South Africa, and the experiments were supervised by local ethical committee members. Animal handling and anesthesia. Following an overnight fast and 2 h without water, the giraffes were premedicated by remote injection (Daninject, Børkop, Denmark) with medetomidine ( 8 g/kg; dose calculated from the estimated body mass). After 5 min, the sedated giraffe was guided to a chute where a halter was mounted on the head, and the animal was blindfolded. Anesthesia was induced with etorphine (3.9 g/kg) and ketamine (0.9 mg/kg im), and after 2 min the giraffe was led into an adjacent pen where it became recumbent within minutes. A rope connected to the halter and passed through a pulley in the ceiling allowed control of the head to avoid injury when the giraffe became recumbent. Immediately after the giraffe became recumbent, a cuffed endotracheal tube (ID 20 mm) was inserted, guided by endoscopy to allow for ventilation with 100% oxygen by using a demand valve (Hudson RCI, Research Triangle Park, NC). The giraffe was then moved to an adjacent room and placed in right lateral recumbency with the head elevated 1 m (Fig. 1A). Electrocardiogram, rectal temperature, end-tidal carbon dioxide tension (PET CO2 ), and tail cuff arterial pressure were recorded using a portable system (model PM9000Vet; E-Vet, Haderslev, Denmark). Blood gas variables were monitored (Opti CCA Apparatus; Osmetech, Roswell, Georgia) from blood collected from an auricular arterial catheter. Anesthesia was maintained by intravenous infusion of -chloralose at 30 mg/kg estimated body mass/h through a catheter in the saphenous vein. Infusion was decreased to 20 mg kg 1 h 1 after 80 min, 15 mg kg 1 h 1 after 120 min, and, thereafter, gradually to 3 mg kg 1 h 1 over 4 5 h depending on reflexes and breathing pattern. Following local infiltration by lidocaine (2%, SAD, Copenhagen, Denmark), vascular sheaths were placed in the carotid artery and jugular vein at the base of the neck and sutured. In addition, two sheaths were placed in the carotid artery and the jugular vein 20 cm below the joint of the jaw and sutured. Straps were placed around each limb leaving the thoracic and abdominal regions free of external pressure, and the giraffe was hoisted to an upright position while the head was supported by a strap connected to the halter (Fig. 1B). This setup allowed the position of the head to be lowered to the level of the heart with a swing to the right or left side (Fig. 1C). It was not possible Fig. 1. A: anesthetized giraffe placed on its right side with the head elevated 1 m by a ladder supported by an oil drum during instrumentation. Sheaths are placed in the upper and lower part of the left carotid artery and jugular vein, while blood pressure is monitored noninvasively. The neck has been shaved to allow ultrasonographic imaging of major vessels. A catheter is positioned in the saphenous vein for infusion of anesthetics. B: the giraffe is suspended in the upright position. C: the head is lowered. to lower the head below heart level because that required active ventroflexion. To avoid influence of the induction agents (medetomidine and etorphine) on the vascular measurements by artificially elevating blood pressures, the effects were antagonized with naltrexone, and measurements were started 90 min following its administration by which time the effects of ketamine would be minimal (32, 40). Hemodynamic variables. Blood pressure was determined by inserting tip transducer catheters through the sheaths with a single pressure sensor mounted on the side of the tip (5 French Micro-Tip SPC 350, range mmhg; Millar Instruments, Houston, TX). Two catheters were inserted through the sheaths at the base of the neck; one in the carotid artery was forwarded retrograde to the aortic arch for recording of MAP, while the catheter in the jugular vein was advanced toward the heart for recording of CVP. In addition, two catheters were inserted through the sheaths at the cranial end of the neck and were directed toward the heart to record the pressure profiles in steps of 20 cm. Carotid artery and jugular venous flows were recorded by transit time probes (Transonic, Ithaca, NY) inserted at the base of the neck. Cross-sectional area and blood velocity of the vessels were determined by ultrasound (Vivid I; GE Healthcare, Horton, Norway) with a 7-MHz linear transducer. Tissue pressure in the upper and lower part of the neck was assessed by positioning subcutaneous fluid-filled catheters (14 gauge; Arrow Teleflex). Spinal fluid pressure was Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

96 R1060 VENOUS POOLING IN THE GIRAFFE NECK Table 1. Morphological characteristics for the studied giraffes Giraffe No * Means SD Body mass, kg Height, cm 322 ND ND Number of valves ND 7 4 Heart weight, kg Heart weight/body weight % *Not included in in vivo measurements because the giraffe died due to fracture of the third cervical vertebra when falling backward following anesthesia. Values are for either the right or left jugular vein. Heart mass and jugular vein length were obtained after the tissue was fixed in 4% formaldehyde. ND, not determined. measured in two giraffes by puncture of the cisterna magna. All measurements were sampled at 100 Hz using data acquisition software (AqKnowledge 3.7.2; Biopack Systems), and an 8-track DAT instrumentation recorder (model D-180; pulse-code modulation, TEAC, Tokyo, Japan) equipped with a microphone to mark events. In two giraffes, carotid arterial and jugular venous blood samples were withdrawn in both the upright position and after the head had been lowered for measurements of oxygen concentration (Osmetech Opti CCA Apparatus). Assuming that oxygen uptake of the giraffe was unaffected by head position, changes in the arterial-venous oxygen concentration difference allowed for an estimate of changes in cardiac output using the Fick equation. Protocol. After instrumentation, the giraffes were suspended in an upright position, and carotid and jugular pressure profiles were obtained with the head in the upright position. Subsequently, hemodynamic variables and vessel cross-sectional areas were obtained while changing the head position by bending the neck away from or toward the monitored vessels for 1 min. All giraffes survived the procedures and were euthanized with intravenous pentobarbital (SAD) and weighed. The heart, carotid arteries, and jugular veins were then obtained. Carotid arteries and the jugular veins were dissected free from connective tissue, and the pericardium, aorta, and pulmonary arteries were removed before the tissue was fixed in 4% formaldehyde. After fixation, the hearts were weighed, and the carotid and the jugular veins were cut open. Telemetric monitoring. One giraffe was successfully instrumented with an implantable telemetric setup (model E-4311; EndoSomatic Tech) connected to pressure catheters (model SPC 350; Millar) for recording of carotid arterial and jugular venous pressures. This giraffe was awake and freely walking in a 4 4-m pen for 48 h before it was resedated, and experiments were performed as described. Statistical analysis. Pressure and flow recordings in the upright and head-lowered positions were compared using one-way ANOVA for repeated measures. A statistical significance level of P 0.05 was used, and data are expressed as means SE. Telemetric measurements are presented as the average of five recordings. RESULTS The morphological characteristics of the giraffes are presented in Table 1. The giraffes were suspended in an upright Table 2. Blood gas variables during experimental period position within 95 6 min after premedication, and the experiments lasted min. Blood pressures increased by 10% during the protocol, and blood gas variables were assessed 5 min after the giraffe became recumbent and four times at predefined time intervals (Table 2). The head was cm above the estimated level of the heart when the giraffes were upright and at about heart level when it was lowered. Histology. The average heart mass was kg, equaling % of body mass (Table 1). Examinations of the jugular veins demonstrated an average of 7 bi- and tricuspid valves, with distance between the valves varying from 2 to 30 cm. One giraffe did not possess any valves in the left jugular vein and only one valve in the right jugular vein. Venous pressure when the head was lowered in that giraffe, however, was not significantly different from that of the others. No valves were observed in the carotid arteries. Head upright. Arterial pressure in the carotid artery was 76 4 mmhg lower at the cranial end than MAP (Table 3). Within the carotid artery, pressure decreased linearly when the catheter was moved distally along the length of the neck, and pressure in the jugular vein was slightly subatmospheric in the cranial end and close to zero at the lower end of the vein (Fig. 2). The lowest venous pressure recorded at the cranial end of the jugular vein was 21 mmhg. Carotid artery blood flow was similar to jugular venous blood flow, and ultrasound showed that the circumference of the carotid artery changed minimally within the cardiac cycle (see Supplemental Movie 1 posted with the online version of this article), whereas the jugular vein was almost collapsed. The carotid flow was pulsatile, while flow was steady in the collapsed jugular vein (Fig. 3). Spinal fluid pressure just below the head was 2 0 mmhg, and neck tissue pressure was generally positive but varied among recording sites. At the cranial and the lower part of the neck, tissue pressures were and mmhg (n 3), respectively. Downloaded from by on May 2, 2017 Minute, position ph PaCO 2, mmhg PaO 2, mmhg HCO 3,mM Base Excess, mm SaO 2,% 5, recumbent , instrumentation , instrumentation , upright , upright Values are means SE, n 5. Variables presented at 5 min after the giraffe became recumbent. Instrumentation at 20 and 60 min is the time of surgical procedures. Upright at 180 and 300 min is the time with the giraffe hoisted to an upright position. Pa CO2, arterial carbon dioxide tension; Pa O2, arterial oxygen tension; HCO 3, arterial bicarbonate; Sa O2, hemoglobin oxygen saturation. AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

97 Table 3. Cardiovascular, spinal fluid, and blood gas variables in the anesthetized giraffe with the head in the upright position and with the head lowered for 1 min to the level of the heart VENOUS POOLING IN THE GIRAFFE NECK R1061 Head Upright Head Lowered Cranial arterial pressure, mmhg * 5 Mean arterial pressure, mmhg * 5 Cranial venous pressure, mmhg * 5 Central venous pressure, mmhg * 5 Spinal fluid pressure, mmhg * 2 Blood velocity, cm/s Carotid flow, l/min Jugular flow, l/min Carotid cross-sectional area, cm Jugular cross-sectional area, cm * 4 Heart rate, beats/min Arterial oxygen tension, mmhg Arterial carbon dioxide tension, mmhg Hemoglobin saturation, % Values are mean SE. No., number of animals studied. *P 0.05; n 1. Carotid arterial and jugular venous pressures measured telemetrically in a freely walking giraffe were and mmhg, respectively. When the same giraffe was anesthetized and suspended in an upright position, telemetrically recorded values were and mmhg, respectively. Heart rate of the freely walking giraffe was 41 1 and 42 0 beats/min when the animal was anesthetized. After all measurements were obtained, the support of one giraffe was moved from the base of the legs to the thoracic and abdominal regions, thereby exerting pressure to these regions. As a result, CVP increased by 14 mmhg, carotid flow decreased, and central venous oxygen saturation decreased to 60% (the detection limit of the apparatus). This indicates that cardiac output fell by 40%. Lowering of the head. When the head was lowered, spontaneous ventilation ceased, and ventilation was then supported. Yet, an increase in Pa CO2 and a 15% reduction in central venous oxygen saturation were observed, while 100% arterial oxygen saturation was maintained (Table 3). Thus, cardiac output decreased by 30%, provided that oxygen uptake did not change. Lowering and raising the head caused substantial changes in neck blood flows and pressures but did not affect heart rate (Table 3). Figure 4 shows a recording during the 1-min lowering of the head to the right side (away from the catheters) and subsequent lifting of the head. At the cranial end of the carotid artery, pressure increased immediately to mmhg when the head was lowered and then gradually decreased to mmhg. This drop in distal carotid arterial pressure was accompanied by a drop in MAP to mmhg (32 7%). Accordingly, with the head lowered, arterial pressure at the base of the skull was similar to MAP (Fig. 4, A and B). The decrease in MAP was paralleled by a decrease in CVP (Fig. 4D) and a 6 mmhg increase in spinal fluid pressure. Carotid blood flow increased immediately when the head was lowered and then returned to the level established when the head was upright (Fig. 4C). Jugular venous flow at the base of the neck ceased for 30 s after the head was lowered and then recovered (Fig. 4F). No. Fig. 2. Pressure profile of the carotid artery and jugular vein in 5 anesthetized giraffes suspended in the upright position. Measurements represent steps of 20 cm (up to 120 cm when length of the neck allowed). Grey lines presents individual values; black line is average SE. Cessation of jugular flow coincided with a progressive rise in cranial jugular pressure (Fig. 4E) and was associated with distension of the vein while no distension was observed in the carotid artery (Fig. 5). With the head lowered, flow in the Fig. 3. Spectral curve of Doppler detection of flow velocity in the cranial part of the carotid artery (top) and the jugular vein (bottom) in an anesthetized giraffe suspended in the upright position. Left: placement of the sample volume using color flow technique. Flow is pulsatile in the carotid artery, while it is steady in the jugular vein. Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

98 R1062 VENOUS POOLING IN THE GIRAFFE NECK A D Fig. 4. Trace of simultaneous venous and arterial measurements of central and cranial pressures (P), and flow in the neck when changing the head position of an anesthetized giraffe suspended in the upright position. Gray bars indicates the time the head was lowered. jugular vein, as detected by Doppler, only occurred when the neck was bent away from the catheters. Lifting the head to the upright position immediately reestablished the pressure difference between the cranial and central part of the carotid artery and caused a large transient rise in jugular blood flow (peak flow 12 l/min) (Fig. 4F). Integration of venous return indicated that l of blood accumulated in the jugular veins when the head was lowered. CVP was restored within 2 4 heart beats after the head was elevated, whereas it took 1 min to normalize MAP. The surge of venous blood returning to the heart when lifting the head coincided with a decrease in the jugular venous cross-sectional area that demonstrated an immediate emptying of the vessel (Fig. 5). Emptying of the vein s distal parts was visualized by ultrasound (see Supplemental Movie 2 at online version of this article). The ultrasound recording also showed that the carotid artery had little or no pulse-mediated change of its diameter during movement of the giraffe s head. DISCUSSION B C This study confirmed the high MAP in giraffes (13 16, 18, 24, 35 37, 39) and that pressure in the carotid artery decreases above heart level in accordance with gravity (18, 37). As a result, the inflow pressure to the brain in the upright position corresponds to that of other mammals, including humans (38). The ultrasound images of the jugular vein revealed a collapsed vein in the upright position and accumulation of venous blood when the head was lowered, which affected MAP. Histological findings. A somewhat surprising histological finding was that the high blood pressure was not correlated with a large heart relative to body mass. The 0.5% relative heart mass found in the six young giraffes studied is similar to that of other mammals, including humans (11). Previous estimates of the relative heart mass of giraffes has been up to 2.3% of body mass (30). That estimate is based on the one report of a 11.3-kg heart from a fully grown giraffe by Goetz and Keen (15). However, body mass was not reported for the giraffe investigated, and it is possible that their reported heart mass included both blood and pericardial fluid. The relative heart mass that we found also agrees with an early report on giraffes by Crisp (9). An interesting implication from this observation is that the generation of the higher pressures takes place without the development of left ventricular hypertrophy that is commonly observed for humans with hypertension (22). The special architecture of the giraffe heart with a thick left ventricular wall, also referred to as gothic (7), might be involved in the heart s increased ability to generate pressure, but the architecture of the heart needs further analysis for detailed evaluation of how a large blood pressure can be generated without myocardial hypertrophy. Histological examinations revealed no valves in the carotid arteries, although such arterial valves have been mentioned in more popular accounts, such as the Encyclopedia Britannica Online. In the jugular vein, no uniform pattern in the distribution of bi- vs. tricuspidal valves was found, and, surprisingly, one giraffe possessed no valves in the (left) jugular vein. Jugular venous flow and pressure. The cerebral circulation of the giraffe has been suggested to be governed by a siphon mechanism where the energy used to overcome gravity in the carotid artery is recovered on the venous side (2, 5). Numerous theoretical (6, 30) and mechanical models (4, 19, 25, 30), however, have not supported this notion and stress that a collapsed vein cannot support a siphon and venous return. Venous return may be better described by a waterfall analogy, where blood seeps through the small lumen of the vein. E F Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

99 VENOUS POOLING IN THE GIRAFFE NECK R1063 Our recordings of venous pressure close to zero throughout most of the length of the jugular vein demonstrate that the jugular vein had a minute cross-sectional area, in agreement with the waterfall analogy. We also observed a maintained small lumen in the cranial part of the jugular vein of the anesthetized giraffe, implying that a column of blood may be preserved in this part of the jugular vein. Consistent with this column of blood, there was recorded negative venous pressure at the cranial end of the jugular vein, while pressures in the dominant part approached zero. The pressures measured in our study agree with McCalden et al. (24), who measured a jugular venous pressure of 0 mmhg at the level of the first cervical vertebrae, while Mitchell and Skinner (28) measured a jugular venous pressure of 10 mmhg some 30 cm below the head. In contrast, Hargens et al. (18) reported that pressure decreased from 13 mmhg at the base of the skull to 4 mmhg 30 cm above heart level in sedated giraffes. It is not evident why jugular venous pressure varies so much between studies. The experiment in which one giraffe was supported under the thoracic region suggests that the protocol used for suspension substantially affects the hemodynamic variables, but the key finding of the present study is that jugular venous pressure of the giraffe depends on the position of the neck. Yet, we acknowledge that differences in the reported jugular venous pressures might also reside from differences in protocols for Fig. 5. Ultrasound image showing the carotid artery (arrow) and the jugular vein (arrowhead) in the cranial part of the neck of an anesthetized giraffe suspended in the upright position (left) and with the head lowered to heart level (right). The jugular vein is distended during lowering of the head, while the carotid artery is unchanged. Bottom: cross-sectional area of the carotid artery and jugular vein in the anesthetized giraffes when the head is upright and when it is lowered. Means SE; *P 0.05 head up vs. lowered (n 4). sedation and anesthesia, although such differences are more likely to affect arterial than venous pressure. Jugular viscous resistance. As pointed out by Badeer (3), pressure at any given location in the jugular vein is the sum of the viscous flow pressure and the gravitational (hydrostatic) pressure. Thus, the low pressure in the jugular vein close to the brain supports a certain venous drag on cerebral blood flow, but along the length of the jugular vein, viscous resistance counterbalances the influence of gravity on pressure. A negative pressure, therefore, implies that a siphon mechanism is active in the upper part of the jugular vein, whereas the pressure 1 m above the heart indicates that the vein is functionally collapsed. From that perspective, the lowest recording of cranial jugular pressure ( 21 mmhg) indicates that the open part of the vein may reach some 30 cm. The collapsed vein was confirmed by ultrasound, and the flow measurements suggest that flow is steady through the small lumen of the collapsed vein. Spinal fluid pressure below the level of the brain was close to zero, in accordance with the estimate for humans (10). By integrating flow, cross-sectional area of the jugular vein, and the blood viscosity for the giraffe (15) in Poiseuilles law, we estimated the relationship between jugular venous crosssectional area and the viscous pressure drop in a 1-m long segment of the vein (Fig. 6). The estimated viscous pressure drop is well within a range where it can counterbalance the Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

100 R1064 VENOUS POOLING IN THE GIRAFFE NECK Fig. 6. Change in viscous pressure drop per meter as a function of jugular venous cross-sectional area for jugular blood flow at 0.6 l/min in the giraffe. Blood viscosity is taken to be kg m 1 s 1 (15). Dotted lines indicate cross-sectional area of the jugular vein as measured by ultrasound in the upright position ( cm 2, n 4; mean SE) and when the head was lowered ( , n 4) in anesthetized giraffes. gravitational pressure increase since the small cross-sectional area of the vein is rigid (Supplemental Movie 1). This calculation indicates that the collapsed part of the jugular vein operates as a Starling resistor, where the viscous resistance explains the small changes in the venous pressure profile along the jugular vein (20). Lowering the head. When lowering the head to drink, the giraffes spread and bend their front legs and, thereby, also lower the level of the heart. Because the giraffes in this study were suspended, the head could only be lowered to heart level. Lowering of the head to this position caused an immediate increase in carotid pressure at the inlet to the brain that corresponded to the change in gravitational force. Thus, it would be expected that the rise in carotid pressure would have been even higher if the head had been lowered to the ground, as when the giraffes drink. The initial steep rise in pressure was followed by a progressive decrease in carotid pressures at the inflow to the brain, as well as in MAP, and (after 45 s) restoration of jugular flow. A consequence of these observations is that the pressure difference between the cranial part of the carotid and jugular vein is decreasing by 20%, while jugular flow is increasing. This scenario indicates a reduction in cerebral vascular resistance with lowering of the head, as suggested by Mitchell et al. (26). This is somewhat surprising and not consistent with precapillary vasoconstriction during lowering of the head as part of an autoregulatory response (24). The one telemetric recording of carotid pressure in a drinking giraffe also recorded a fall in MAP during drinking (36), but this was accompanied by bradycardia, presumably as a result of stimulation of carotid arterial baroreceptors (21). In the anesthetized giraffes, heart rate did not change when the head was lowered. This could reflect a depression of autonomic reflexes by the anesthesia. Furthermore, the 30% reduction of MAP after lowering the head coincided with a reduction in CVP, both indicating reduced cardiac filling and stroke volume and, hence, cardiac output. The percentile reduction in MAP matches the calculated 30% reduction in cardiac output estimated from the oxygen saturation. This could indicate that the reduction in MAP was independent of a reduction of total peripheral resistance, as expected if the autonomic reflexes are depressed in the anaesthetized giraffes. Although -chloralose, the maintenance drug in our study, provides stable hemodynamic variables over long periods in a wide range of mammals, including domestic ruminants (8, 12, 34), we acknowledge that baro- and volume receptor function might have been affected. However, neither can we exclude the possibility that the baroand volume receptors have sensed the 4% reduction in blood volume. It, therefore, remains to be determined whether changes in overall peripheral resistance contributes to the reduction of MAP when the head is lowered in awake giraffes. The ultrasound images revealed a pronounced expansion of the jugular vein when the head was lowered, and the volume of blood that accumulated in the veins was determined to be 1.2 liters, or 4% of the estimated blood volume (16). The volume in the jugular vein and the increase in spinal fluid pressure may well have been larger if a lower head position had been obtained, as by the drinking position. Nevertheless, the accumulated volume is considerably smaller than previously suggested (26, 27), and it remains to be established whether translocation of this relatively small volume of blood to the jugular vein is sufficient to significantly decrease cardiac filling when the head is lowered. Alternatively, lowering of the head may reduce cardiac filling by altering the hydrostatic indifference point, and it is possible that blood pressure regulation by the arterial baroreceptors positioned at the base of the skull in the giraffe (21) are important. Although these mechanisms warrant further evaluation, we suggest that the reduction in MAP decreased the rise in blood pressure at the level of the head and, thereby, contributed to protecting brain capillaries when the head is lowered. Furthermore, subdural venous collapse (17), along with the increase in spinal fluid pressure when lowering the head, may act to protect the cerebral vasculature by decreasing the pressure difference between the capillary blood and spinal fluid. Lifting of the head. When the head is returned to the upright position, enhanced cardiac filling may be important for maintaining cerebral blood flow and, thereby, preventing (pre)syncopal symptoms. The sudden emptying of the jugular veins when the head was lifted was associated with an increase in carotid blood flow. While the return of venous blood from the neck to the heart took place within seconds, it took more than a minute after the head had been lifted before MAP was reestablished (Fig. 4). Whether the slow return of MAP could be explained by the volume of translocated blood or is a consequence of lacking baroreceptor function is not known. Furthermore, the muscular structure in the lower part of the vena cava (15) or pooling of blood in the pulmonary circulation due to altered breathing patterns in the suspended position could be involved. To evaluate these possibilities requires a setup that allows for measurements in naturally drinking, awake giraffes. Perspectives and Significance In line with the August Krogh principle that suggests that for many physiological problems there will be an animal of choice on which it can be most conveniently studied, giraffes and other long-necked animals can provide fundamental insights to the effects of gravity on the circulatory system of all animals. Simultaneous measurements of blood flow and pressure show that cerebral perfusion of anaesthetized upright positioned giraffes is maintained by the high arterial blood pressure and that the jugular vein is collapsed in the upright position. This implies that the vascular resistance of the jugular vein Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

101 counterbalances the gravitational pressure profile and, thus, there is no indication of a functional siphon. The pronounced accumulation of blood in the jugular vein when the head was lowered affected MAP. This points to an important role of venous distensibility in influencing how gravitational pressures affect the cardiovascular system during postural changes. It is, therefore, interesting to study whether similar mechanisms exist in other long-necked animals such as ostriches and camels, as well as to repeat this study on free-ranging animals (1) to avoid any effect of anesthesia and to obtain measurements from naturally behaving unstressed animals. ACKNOWLEDGMENTS We thank Ismail Laher, Paul R. Manger, Geoffrey Candy, Vinny Naidoo, Wouter van Hoven, Tanja Thomsen, Kirsten Skaarup, Einer Larsen, Jørgen Andresen, and Kamilla Wang for valuable contributions. Charles van Niekerk, Martin Krogh, and the staff at Wildlife Assignments International are acknowledged for expert animal handling during the study. We are grateful to Professor James W. Hicks for discussions of the experiments. GRANTS This study was supported by the Lundbeck Foundation, The Danish Heart Association, the Aase and Ejnar Danielsen Foundation, The Danish Research Council, and the Faculties of Health Science and Natural Sciences at University of Aarhus, Denmark. DISCLOSURES GE Healthcare Denmark provided ultrasound apparatus and Danish Myo Technology, Denmark provided technical assistance. REFERENCES 1. Axelsson M, Dang Q, Pitsillides K, Munns S, Hicks J, Kassab GS. A novel, fully implantable, multichannel biotelemetry system for measurement of blood flow, pressure, ECG, and temperature. J Appl Physiol 102: , Badeer HS. Does gravitational pressure of blood hinder flow to the brain of the giraffe? Comp Biochem Physiol A 83: , Badeer HS. Haemodynamics of the jugular vein in the giraffe. Nature 332: , Badeer HS. Is the flow in the giraffe s jugular vein a free fall? Comp Biochem Physiol A 118: , Badeer HS, Hicks JW. Hemodynamics of vascular waterfall is the analogy justified? Respir Physiol 87: , Brook BS, Pedley TJ. A model for time-dependent flow in (giraffe jugular) veins: uniform tube properties. J Biomech 35: , Coghlan C, Hoffman J. Leonardo da Vinci s flights of the mind must continue: cardiac architecture and the fundamental relation of form and function revisited. Eur J Cardiothorac Surg 29, Suppl 1: S4 S17, Constable PD. Acute endotoxemia increases left ventricular contractility and diastolic stiffness in calves. Shock 12: , Crisp E. Contributions to the anatomy of the giraffe, with an account of the length of the alimentary canal of many of the ruminants, as measured by the author. Proc Zool Soc London 63 68, Dawson EA, Secher NH, Dalsgaard MK, Ogoh S, Yoshiga CC, Gonzalez-Alonso J, Steensberg A, Raven PB. Standing up to the challenge of standing: a siphon does not support cerebral blood flow in humans. Am J Physiol Regul Integr Comp Physiol 287: R911 R914, de la Grandmaison GL, Clairand I, Durigon M. Organ weight in 684 adult autopsies: new tables for a caucasoid population. Forensic Sci Int 119: , Fauchere JC, Walker AM, Grant DA. Right atrial pressure as a measure of ventricular constraint arising from positive end-expiratory pressure during mechanical ventilation of the neonatal lamb. Crit Care Med 31: , Goetz RH. The giraffe. Lancet 271: 351, VENOUS POOLING IN THE GIRAFFE NECK R Goetz RH, Budtz-Olsen O. Scientific safari the circulation of the giraffe. S A Med J , Goetz RH, Keen EN. Some aspects of the cardiovascular system in the giraffe. Angiology 8: , Goetz RH, Warren JV, Gauer OH, Patterson JL, Doyle JT, Keen EN, Mcgregor M. Circulation of the giraffe. Circ Res 8: , Grande PO. Mechanisms behind postspinal headache and brain stem compression following lumbar dural puncture a physiological approach. Acta Anaesthesiol Scand 49: , Hargens AR, Millard RW, Pettersson K, Johansen K. Gravitational hemodynamics and edema prevention in the giraffe. Nature 329: 59 60, Hicks JW, Badeer HS. Siphon mechanism in collapsible tubes: application to circulation of the giraffe head. Am J Physiol Regul Integr Comp Physiol 256: R567 R571, Holt JP. The collapse factor in the measurement of venous pressure: the flow of fluid through collapsible tubes. Am J Physiol 134: , Kimani JK, Mungai JM. Observations on the structure and innervation of the presumptive carotid-sinus area in the giraffe (Giraffa camelopardalis). Acta Anat (Basel) 115: , Lekakis JP, Zakopoulos NA, Protogerou AD, Kotsis VT, Papaioannou TG, Stamatelopoulos KS, Tsitsiricos MD, Pitiriga VC, Papamichael CM, Toumanides ST, Mavrikakis ME. Cardiac hypertrophy in hypertension: relation to 24-h blood pressure profile and arterial stiffness. Int J Cardiol 97: 29 33, Lyon CK, Scott JB, Wang CY. Flow through collapsible tubes at low Reynolds numbers. Applicability of the waterfall model. Circ Res 47: 68 73, McCalden TA, Borsook D, Mendelow AD, Shimell CJ, Vos VD, Pieterse PC, Klerk BD. Autoregulation and hemodynamics of giraffe carotid blood-flow. S Afr J Sci 73: , Mitchell G, Bobbitt JP, Devries S. Cerebral perfusion pressure in giraffe: modelling the effects of head-raising and -lowering. J Theor Biol 252: , Mitchell G, Hepburn HR, Mitchell D, Shimell C, Borsook D, Ford DM, Klugman K, Rogers G. Carotid hemodynamics in the giraffe (Abstract). S Afr J Sci 76: 246, Mitchell G, Maloney SK, Mitchell D, Keegan DJ. The origin of mean arterial and jugular venous blood pressures in giraffes. J Exp Biol 209: , Mitchell G, Skinner JD. How giraffe adapt to their extraordinary shape. Trans R Soc S Afr 48: , Paulson OB, Strandgaard S, Edvinsson L. Cerebral autoregulation. Cerebrovasc Brain Metab Rev 2: , Pedley TJ, Brook BS, Seymour RS. Blood pressure and flow rate in the giraffe jugular vein. Philos Trans R Soc Lond B Biol Sci 351: , Pott F, Van Lieshout JJ, Ide K, Madsen P, Secher NH. Middle cerebral artery blood velocity during intense static exercise is dominated by a Valsalva maneuver. J Appl Physiol 94: , Ranheim B, Arnemo JM, Ryeng KA, Soli NE, Horsberg TE. A pharmacokinetic study including some relevant clinical effect of medetomidine and atipamezole in lactating dairy cows. J Vet Pharmacol Ther 22: , Seymour RS. Model analogues in the study of cephalic circulation. Comp Biochem Physiol A Mol Integr Physiol 125: , Silverman J, Muir WW III. A review of laboratory animal anesthesia with chloral hydrate and chloralose. Lab Anim Sci 43: , Van Citters RL, Franklin DL, Vatner SF, Patrick T, Warren JV. Cerebral hemodynamics in the giraffe. Trans Assoc Am Physicians 82: , Van Citters RL, Kemper WS, Franklin DL. Blood pressure responses of wild giraffes studied by radio telemetry. Science 152: , Van Citters RL, Kemper WS, Franklin DL. Blood flow and pressure in the giraffe carotid artery. Comp Biochem Physiol 24: , Van Lieshout JJ, Wieling W, Karemaker JM, Secher NH. Syncope, cerebral perfusion, oxygenation. J Appl Physiol 94: , Warren JV. The physiology of the giraffe. Sci Am 231: , Waterman AE. The pharmacokinetics of ketamine administered intravenously in calves and the modifying effect of premedication with xylazine hydrochloride. J Vet Pharmacol Ther 7: , Downloaded from by on May 2, 2017 AJP-Regul Integr Comp Physiol VOL 297 OCTOBER

102 Sports Med (2015) 45 (Suppl 1):S23 S32 DOI /s REVIEW ARTICLE The Physiological Regulation of Skeletal Muscle Fatty Acid Supply and Oxidation During Moderate-Intensity Exercise Gerrit van Hall 1 Published online: 9 November 2015 The Author(s) This article is published with open access at Springerlink.com Abstract Energy substrates that are important to the working muscle at moderate intensities are the non-esterified fatty acids (NEFAs) taken up from the circulation and NEFAs originating from lipolysis of the intramuscular triacylglycerol (IMTAG). Moreover, NEFA from lipolysis via lipoprotein lipase (LPL) in the muscle of the very-lowdensity lipoproteins and in the (semi) post-prandial state chylomicrons may also contribute. In this review, the NEFA fluxes and oxidation by skeletal muscle during prolonged moderate-intensity exercise are described in terms of the integration of physiological systems. Steps involved in the regulation of the active muscle NEFA uptake include (1) increased energy demand; (2) delivery of NEFA to the muscle; (3) transport of NEFA into the muscle by NEFA transporters; and (4) activation of the NEFAs and either oxidation or re-esterification into IMTAG. The increased metabolic demand of the exercising muscle is the main driving force for all physiological regulatory processes. It elicits functional hyperemia, increasing the recruitment of capillaries and muscle blood flow resulting in increased NEFA delivery and accessibility to NEFA transporters and LPL. It also releases epinephrine that augments adipose tissue NEFA release and thereby NEFA delivery to the active muscle. Moreover, NEFA transporters translocate to the plasma membrane, further increasing the NEFA uptake. The majority of the NEFAs taken up by the active muscle is oxidized and a minor & Gerrit van Hall Gerrit.van.hall@regionh.dk 1 Clinical Metabolomics Core Facility, Department of Biomedical Sciences, Faculty of Health and Medical Sciences, Rigshospitalet, University of Copenhagen, Section 7652, 9 Blegdamsvej, 2100 Copenhagen, Denmark portion is re-esterified to IMTAG. Net IMTAG lipolysis occurs; however, the IMTAG contribution to total fat oxidation is rather limited compared to plasma-derived NEFA oxidation, suggesting a complex role and regulation of IMTAG utilization. 1 Introduction Limitation of carbohydrate and lipid transfer from the microvascular system to the muscle cell occurs at the onset of exercise when the delivery and transport systems are not optimal and during continuous exercise above moderate intensities of 50 % of maximal pulmonary oxygen uptake (VO 2max ). At higher continuous workloads the extracellular substrate provision rate is not high enough and intracellular stored substrates must also be used, such as glycogen and intramuscular triacylglycerol (IMTAG) [1]. Due to complexity in the regulation of fat metabolism, it is still unknown what limits active muscle fat oxidation [2 4]. Various suggestions have been put forward, from limitation in the non-esterified fatty acids (NEFAs) delivery to the active muscle, to NEFA uptake into muscle or into the mitochondria or b-oxidation. In addition, the IMTAG not limited by delivery or uptake seems to be not utilized optimally as its pool is not even close to fully utilized [5]. The plasma NEFA concentration plays an important role in the NEFA uptake and subsequent oxidation by the active muscle. If the plasma NEFA concentration falls, the rate of muscle NEFA uptake and subsequent oxidation will decline as well. Conversely, an increase in the plasma NEFA concentration will increase the active muscle NEFA uptake and oxidation [6], similar to blood glucose and its uptake by the active muscle [7]. Adipose tissue release of NEFAs is the primary source of NEFAs by which the 123

103 S24 plasma NEFA concentration is sustained under post-absorptive conditions or increased during exercise [8]. Thus, the control of adipose tissue triacylglycerol (TAG) lipolysis and subsequent release of the NEFAs into the circulation has an important role in the regulation of muscle NEFA uptake and oxidation during exercise. The liver also plays an important role as it has a high NEFA uptake and, thereby, can affect the plasma NEFA concentration, albeit changes in liver NEFA uptake during exercise are not welldescribed [9]. Moreover, the NEFA delivery to the active muscles (NEFA concentration 9 plasma flow) is an even better determinant of the NEFA uptake than the NEFA concentration [10]. The blood flow to the active muscles increases several-fold upon exercise, primarily to increase oxygen supply, but it will also increase NEFA delivery. NEFA uptake from the plasma into the cytosol may occur to some extent via passive diffusion but over the past decades it has been shown that the majority of NEFA uptake occurs via facilitated transport and that muscle contraction induces plasma membrane fatty acid translocase (FAT/CD36) and fatty acid binding protein (FABPpm) translocation from the intracellular depots to the plasma membrane [11]. Therefore, NEFA transport into the muscle cells may be a potential regulatory and limiting step in NEFA utilization by muscle during exercise. Once in the cytosol, the NEFA is activated and can either be oxidized or stored in IMTAG. In this review, the control of human in vivo NEFA fluxes and subsequent oxidation by skeletal muscle during prolonged moderate-intensity exercise are described in terms of the integration of physiological systems (Fig. 1). 2 Skeletal Muscle Non-Esterified Fatty Acid (NEFA) Oxidation During Exercise from Plasma-Derived NEFA and Intramuscular Triacylglycerol Lipolysis The regulation of the active muscle fatty acid uptake and subsequent oxidation is complex and may be differently regulated at low, moderate, and high exercise intensities and durations of exercise. Relatively few studies have directly determined human muscle NEFA handling during exercise, with most of the outstanding studies performed in the s [12 15], and even fewer studies have determined IMTAG involvement [6]. The limited, but remarkably consistent, available quantitative and kinetic information originates from prolonged moderate-intensity exercise. The regulation of the active muscle NEFA uptake can be defined by a four-step process, as depicted in Fig. 1, consisting of (1) an increased energy demand by the contracting muscle; (2) delivery of NEFA to the muscle; (3) transport of NEFA into the muscle by fatty acids G. van Hall transporters; and (4) activation of the fatty acids and either oxidation or re-esterified into intracellular lipids and stored into the IMTAG droplets located next to the mitochondria. A similar concept is generally acknowledged for glucose [7, 16] but differences seem to exist with respect to glycogen versus IMTAG as the intramuscular carbohydrate and fat energy stores, respectively. The increased metabolic rate/demand of the active muscle is usually not considered to be a regulatory step. However, it is the main driving force for all physiological regulatory processes. It elicits functional hyperemia, increasing muscle blood flow, the number of perfused capillaries (recruitment), and hormone levels that affect adipose tissue NEFA release, and hence NEFA delivery to the active muscle. Moreover, resting skeletal muscle has very low energy expenditure and thus the demand for energy/adenosine triphosphate is small, implying that an increase in energy demand with exercise causes a surge for substrates that possibly creates a concentration gradient for NEFA between plasma, the interstitial space, cytosol, and entry in the mitochondria (Figs. 1, 2). The resting muscle blood flow is low and % of all NEFA delivered to the muscle is taken up [10, 12, 15, 17] and is clearly dependent on the NEFA concentration [6]. This suggests that the facilitated muscle NEFA uptake is limited/saturable or that the lack of substrate demand and/ or re-esterification into intracellular lipids is low, reducing the NEFA gradient for uptake. During moderate-intensity exercise the blood flow to the active muscles increases linearly with the workload [18] and is easily 10- to 15-fold higher during moderate-intensity exercise [17] than at rest, and with the increase in blood flow the blood transit time through the active muscle is substantially reduced [19, 20]. The NEFA fractional extraction decreases to only *20 % [6, 10, 12 15, 17] as compared to the % at rest, despite the massive increase in blood flow and reduced transit time. In addition, the NEFA factional extraction is the same over a wide range of plasma flows [12], i.e., exercise intensities, and is independent of the NEFA concentration over at least a threefold increase in NEFA concentration [6, 15]. Accordingly, the active muscle NEFA uptake is very closely and linearly related to the NEFA delivery, which is the NEFA concentration multiplied by the plasma flow to the active muscle [6, 10, 12 15, 17]. The same relationship is found between NEFA delivery and NEFA oxidation, since % of the NEFA taken up by the active muscle is directly oxidized [6, 10, 12 14, 17]. Thus, during exercise the increase in NEFA uptake with delivery is dependent on the functional exercise hyperemia increase in blood flow. However, the linear increase of NEFA delivery and oxidation with the duration of prolonged moderate-intensity exercise is caused by the increase in NEFA concentration since the 123

104 Muscle Fatty Acid Supply and Oxidation During Prolonged Exercise S25 Adipose ssue NEFA release TAG lipolysis Blood flow Capillary recruitment GI tract Food/fat inges on CM-TAG release? Muscle NEFA delivery NEFA concentra on Blood flow Capillary recruitment?? Liver NEFA uptake NEFA oxida on Ketone body produc on VLDL-TAG produc on NEFA transport Transloca on FAT/CD36/FABPpm NEFA transporters Ac va on acyl-coa synthetase NET IMTAG lipolysis NEFA IMTAG synthesis IMTAG lipolysis NEFA oxida on Energy demand (ATP) Ac ve muscle Fig. 1 Schematic representation of the control of non-esterified fatty acid (NEFA) fluxes and oxidation during exercise described in terms of the integration of physiological systems. Central in the scheme is the close linear relationship between the active muscle NEFA delivery and uptake/oxidation observed during continuous moderate-intensity exercise. The increase in NEFA delivery is caused by an increase in plasma flow that includes capillary recruitment, whereby a larger portion of the total available NEFA transport machinery becomes accessible. Plasma flow and capillary recruitment do not change much with continuous exercise at the same workload; hence, the increase in muscle NEFA delivery and uptake/oxidation with exercise duration is mainly mediated by the increase in NEFA concentration. The increase in the plasma NEFA concentration is facilitated by an increased release of NEFA from adipose tissue via increased adipose tissue triacylglycerol (TAG) lipolysis, adipose tissue blood flow, and capillary recruitment with epinephrine and possible atrial natriuretic peptide as the key regulators during exercise. Liver NEFA uptake at rest and during exercise is substantial but does not seem to change much with exercise, thus it does not have much effect on the plasma NEFA concentration. Mandatory, but likely not limiting for NEFA oxidation in healthy individuals, is the facilitated NEFA uptake by transporters and binding proteins of which fatty acid translocase (FAT/CD36) and fatty acid bounding protein (FABPpm) translocate from the intracellular depot(s) to the plasma membrane with muscle contraction. NEFA from intramuscular TAG (IMTAG) is used during moderate-intensity exercise. The net degradation of IMTAG is caused by a decrease in IMTAG synthesis and maintained or increased lipolysis. Moreover, NEFA originating from either very-low-density lipoproteins (VLDL-TAG), or chylomicrons (CM-TAG) primarily in the fed state, may contribute to the active skeletal muscle NEFA oxidation. Dietary fat reaches the liver, adipose tissue, and skeletal muscle in the form of CM-TAG that undergoes lipolysis by lipoprotein lipase (LPL) and the resulting NEFAs are taken up and either oxidized or esterified by the active muscle (see also Fig. 2). The contribution of VLDL-TAG- and CM- TAG-derived NEFA to the resting and active muscle energy requirements seems limited. ATP adenosine triphosphate, CoA coenzyme A, GI gastrointestinal plasma flow is largely unchanged during continuous exercise at the same intensity, emphasizing the important role of an increased adipose tissue NEFA release to enhance the plasma NEFA concentration driving the active muscle NEFA oxidation (Fig. 1 and Sect. 3) [8]. Consistent with these findings, an increase in the NEFA concentration via intralipid/heparin infusion increases the active muscle net NEFA uptake and fat oxidation [21], and even total fat oxidation at the relative high workload of 85 % of VO 2max [22]. Conversely, if NEFA concentrations are lowered via nicotinic acid infusion, the plasma NEFA oxidation is reduced [23]. The capacity for muscle to achieve such a rapid and manyfold increase in NEFA uptake upon contraction is remarkable, particularly in view of the manyfold increase in NEFA delivery caused by the massive increase in blood flow and the substantially reduced time available for interaction with the NEFA transporters. Of course, the 10- to 15-fold increase in blood flow with moderate-intensity exercise is accompanied by an increased capillary recruitment [19, 20], also referred to as nutritive flow [24], increasing the accessibility to NEFA transporters and diffusion surface. In addition, contraction-induced translocation of NEFA transporters from the intracellular storage pool to the cell membrane [11, 25] plays an important role in enhancing muscle NEFA uptake capacity during exercise. Another contributing factor is likely the high energy demand of the active muscle creating a NEFA surge and increasing the gradient for NEFA movement from the mitochondrion to the cytosol and eventual plasma NEFA uptake. 123

105 S26 G. van Hall Glucose Albumin NEFA TAG 1,2-acylglycerol Triacylglycerol ATGL/HSL Phosphatidic acid phosphatase DGAT Diacylglycerol HSL IMTAG Phosphatidic acid Acyl-CoA NEFA Capillary recruitment 1-acylglycerol-3-P Glycerol-3-P Glycerol kinase Glycerol MAGL Monoacylglycerol TAG 1-acylglycerol-3-P acyltransferase GPAT Albumin NEFA FABPpm FAT Transloca on LPL FAT DELIVERY NEFA Acyl-CoA FAT FABPc ACBP Acyl-CoA Glucose GLUT4 Glucose Glucose-6-P Pyruvate Acetyl-CoA Transloca on GLUT4 Phosphorylase Glycogen granules Glycogen synthase Glycogen -oxidation TCA-cycle Myocyte Mitochondrion CO 2 Fig. 2 Schematic representation of skeletal muscle energy metabolism. Two pathways in skeletal muscle energy oxidation during exercise can be recognized: the extracellular and intracellular substrate supply. The increase in the extracellular muscle energy supply during exercise is mediated via an increase in the blood substrate delivery of glucose from either carbohydrate intake or liver glycogenolysis and gluconeogenesis, non-esterified fatty acids (NEFA) mainly from adipose tissue, and chylomicron or very-lowdensity lipoproteins referred to as triacylglycerol (TAG) (see Fig. 1). The increase in delivery of these substrates to the active muscle is mediated by an increase in blood flow, including an increase in capillary recruitment, and substrate concentration. Transport of blood glucose into skeletal muscle is facilitated by glucose transporter-4 (GLUT4) and the long-chain NEFAs via fatty acid transporters (FAT), that also facilitates the transport of NEFA into the mitochondria. The fate of the glucose and NEFA taken up by skeletal muscle is oxidation or storage into glycogen or TAG, respectively. The intracellular energy supply during exercise is immediately increased, mainly via a fast breakdown of glycogen crucial to cover the instantaneously manyfold increase in energy demand going from rest to exercise. The rate of glycogen breakdown decreases with exercise duration, and glucose uptake and subsequent oxidation and later fatty acid oxidation increases. The increase in NEFA availability from intramuscular triacylglycerol (IMTAG) breakdown during exercise is mediated by a reduction in NEFA re-esterification and possibly an increase in IMTAG lipolysis. The role and regulation of the muscle IMTAG turnover rate is unknown. ACBP acyl-coa binding protein, ATGL adipose tissue triglyceride lipase, CoA coenzyme A, DGAT diacylglycerol acyltransferase, FABP fatty acid binding protein (pm plasma membrane, c cytosolic), FABPm fatty acid binding protein, GPAT glycerol-3-phosphate acyltransferase, HSL hormone sensitive lipase, LPL lipoprotein lipase, MAGL monoacylglycerol lipase, TCA tricarboxylic acid Over recent decades there has been considerable debate regarding whether NEFA transport across the plasma membrane occurs via passive diffusion or via facilitated transport by means of membrane-associated proteins. The physical properties of NEFA with a non-polar carbon chain and the polar head group would make passive diffusion possible, and this is seemingly supported by the abovedescribed connection between NEFA concentration and active muscle NEFA uptake. However, over the past decade it has been shown that the majority of NEFA is transported across the cell membrane via protein-mediated mechanisms. Moreover, some of these fatty acid transporters demonstrate reversible translocation, from the intracellular storage pool to the cell membrane [11, 25] under the influence of contraction and insulin, analogous to the well-described glucose transporter type-4 (GLUT-4) 123