Muscle chemoreflex alters carotid sinus baroreflex response in humans.

The arterial baroreflex opposes pressor responses to muscle ischemia (muscle chemoreflex). Our experiments sought to quantify the unknown effects of muscle chemoreflex on carotid sinus baroreflex (CSB) sensitivity. We generated CSB stimulus-response (S-R) curves by pulsatile application (triggered by each electrocardiogram R wave) of positive and negative neck pressure (from 60 to -80 mmHg in 20-mmHg steps of 20 s each) in seven normal young men. S-R curves were obtained at rest (upright), during the last 3 min of upright cycle ergometer exercise (150 W), and at the first minute of postexercise recovery with leg circulation free (control). A second study repeated the same procedures, except that leg circulation was occluded 20 s before the end of exercise to elicit muscle chemoreflex, and occlusion was maintained during recovery measurements (approximately 3- to 4-min duration). S-R curves for CSB were shifted upward and rightward (25 mmHg) to higher arterial blood pressure (BP) by exercise and less so (10 mmHg) in recovery (free leg flow). Postexercise occlusion (muscle chemoreflex) raised BP and shifted S-R curves above exercise curves. CSB gain rose from -0.26 +/- 0.06 (control) to -0.44 +/- 0.08 (occlusion) during positive neck pressure application and was reduced from -0.14 +/- 0.04 to zero (-0.04 +/- 0.03) during negative neck pressure. Heart rate responses during postexercise muscle chemoreflex were not significantly different from control. Results reveal a nonlinear summation of CSB and muscle chemoreflex effects on BP. BP-raising capability of muscle chemoreflex enhances CSB responses to hypotension but overpowers baroreflex opposition to hypertension.

Neural control of muscle blood flow: importance during dynamic exercise.

1. The present review examines the control of muscle vascular conductance by the sympathetic nervous system during exercise. 2. Evidence for tonic sympathetic neural control of active muscle rests on three findings: (i) directly measured muscle sympathetic nerve traffic is increased; (ii) spillover of noradrenaline from active muscles is also increased; and (iii) withdrawal of sympathetic outflow to active muscle either by acute blockade of its sympathetic nerve supply or by reflex inhibition of sympathetic nervous activity raises muscle vascular conductance via inhibition of tonic vasoconstriction. 3. Loss of tonic sympathetic control of muscle vascular conductance during mild to severe exercise caused marked hypotension despite maintenance of a normal cardiac output. 4. The extent to which active muscle can vasodilate in intact animals appears to have been hidden by tonic vasoconstriction. This vasoconstriction appears to be minimally affected by metabolites in oxidative (red) muscle, but may be inhibited in predominantly glycolytic (white) muscle owing to different spatial distributions of alpha 1- and alpha 2-adrenoceptors in the two muscle types and to the different susceptibilities of the two receptor types to interference by metabolites. 5. The reflexes causing vasoconstriction in active and inactive muscles are unknown. One hypothesis is that a flow-sensitive muscle chemoreflex raises sympathetic outflow to reduce accumulations of muscle metabolites caused by mismatches between muscle blood flow and metabolism, called 'flow errors'. Another hypothesis is that the arterial baroreflex corrects mismatches between cardiac output and vascular conductance called 'pressure errors'. This review argues for a dominance of control by the baroreflex based on the following observations: (i) the arterial baroreflex is essential to the normal rise in sympathetic nervous activity and arterial pressure at the onset of exercise; (ii) during submaximal exercise, a functioning arterial baroreflex is required to maintain tonic sympathetic activity and prevent arterial hypotension; and (iii) whereas a muscle chemoreflex may be needed to guard against hypoperfusion of active muscle, the arterial baroreflex must oppose hypotension by initiating sympathetic vasoconstriction to oppose muscle vasodilation.

Carotid baroreflex control of blood pressure and heart rate in men during dynamic exercise.

The degree of control of blood pressure (BP) and heart rate (HR) by arterial baroreflex during exercise is still controversial. We studied baroreflex control of BP and HR in seven normal young men by a noninvasive procedure employing a neck suction chamber that delivers pulsatile positive and negative pressures to the carotid sinus (CS). Pressures applied to the CS ranged from -80 to +60 Torr in steps of 20 Torr. Pressure stimuli were triggered by electrocardiogram R wave, and each pressure step was maintained for 20 s in a continuous sequence. One baroreflex-response curve was obtained during the last 3 min of each 6-min period of exercise. The four levels of upright (cycle) exercise were 60, 120, 180, and 240 W, the highest requiring approximately 75% of maximal O2 uptake. The sensitivity of the HR baroreflex response assessed by linear regression of HR vs. CS pressure (CSP) did not significantly decrease from rest (-0.09 +/- 0.053 beat/Torr) to 240 W (-0.06 +/- 0.025 beat/Torr). The BP above or below which CSP was increased or decreased by neck collar pressure was significantly increased from rest (76 +/- 6.5 Torr) to 240 W (111.2 +/- 4.0 Torr). The sensitivity of baroreflex response was assessed by linear regression of BP vs. CSP and was not significantly different from rest (-0.29 +/- 0.054 Torr/Torr) up to exercise at 240 W (-0.29 +/- 0.048 Torr/Torr). We conclude that mild to severe exercise does not reduce the gain of the CS reflex below resting values.(ABSTRACT TRUNCATED AT 250 WORDS)

Is rapid rise in vascular conductance at onset of dynamic exercise due to muscle pump?

We tested the hypothesis that rapid increases in muscle blood flow and vascular conductance (C) at onset of dynamic exercise are caused by the muscle pump. We measured arterial (AP) and central venous pressure (CVP) in nine awake dogs, eight with atrioventricular block, pacemakers, and ascending aortic flow probes for control of cardiac output (CO) (2 also had terminal aortic flow probes). One dog had only an iliac artery probe. At exercise onset (0 and 10% grade, 4 mph) C and CVP rose to early plateaus, and AP reached a nadir, all in 2-5 s. At 20% grade and 4 mph, C increased continuously after its initial sudden rise. Timing and magnitude of initial change in conductance (delta C) were independent of CO, AP, work rate (change in grade at constant speed), or autonomic function (blocked by hexamethonium). Speed of initial delta C and its independence from work rate and blood flow ruled out metabolic vasodilation as its cause; insensitivity to AP and autonomic blockade ruled out myogenic relaxation and sympathetic vasodilation as causes of sudden delta C. Sensitivity to contraction frequency (not work per se) implicates the muscle pump. When reflexes were blocked, a large secondary rise in C, presumably caused by metabolic vasodilation, began after 10 s of mild exercise. When reflexes were intact in mild exercise, C was lowered below its initial plateau by sympathetic vasoconstriction, which partially raised AP from its nadir toward its preexercise level. Our conclusion is that dynamic exercise has a large rapid effect on C that is not explained by known neural, metabolic, myogenic, or hydrostatic influences.(ABSTRACT TRUNCATED AT 250 WORDS)

Dependence of cardiac filling pressure on cardiac output during rest and dynamic exercise in dogs.

At rest, central venous pressure (CVP) falls when cardiac output (CO) rises. This can be attributed to flow-dependent redistribution of blood volume from central to peripheral blood vessels. In contrast, CVP rises during dynamic exercise despite a rise in CO. Therefore peripheral circulatory changes during exercise must counteract the factors that lower CVP when CO rises during rest. Our objectives were to determine the importance of blood flow, the muscle pump, and reflexes on changes in ventricular filling pressure during dynamic exercise. In seven dogs with a surgically produced atrioventricular (AV) block, normal relationships between CO and CVP were established by AV-linked pacing (normal heart rates) during rest and exercise. Cardiac output was altered during rest and treadmill exercise (4 miles/h at 0, 10, or 20% grade) by changing ventricular pacing rate to establish curves relating delta CVP to delta CO. These curves were displaced rightward (higher CO) and upward (higher CVP) by exercise because of the muscle pump. Changing CO by pacing during rest and exercise revealed a constant slope for delta CVP/delta CO of -2.7 mmHg.l-1.min-1. Blockade of reflex vasoconstriction and venoconstriction with hexamethonium at rest and during mild exercise (to isolate effects of the muscle pump) did not alter these slopes or the displacement of the curves by exercise, although CVP was 4.3 mmHg lower at a given CO after blockade.(ABSTRACT TRUNCATED AT 250 WORDS)

Cardiopulmonary and carotid baroreflex control of splanchnic and forearm circulations.

The objective was to determine whether a rise in carotid sinus transmural pressure by neck suction (NS) would counteract vasoconstriction secondary to inhibition of discharge of arterial and cardiopulmonary baroreceptors by simultaneous lower body negative pressure (LBNP). NS alone was applied to seven normal human subjects at -40 mmHg for 400-600 ms at each heartbeat during a 6-min period. NS reduced mean arterial pressure (MAP) from 94 +/- 6 to 86 +/- 9 mmHg and heart rate (HR) from 64 +/- 5 to 60 +/- 4.7 beats/min but did not affect vascular resistance in the splanchnic region (flow by constant infusion of indocyanine green; assumed constant extraction) or in the forearm (venous occlusion plethysmography). The same NS stimulus was applied during 23 min of continuous LBNP at -40 mmHg. LBNP alone before NS significantly reduced central venous pressure (CVP) from 5 +/- 0.3 to 1 +/- 0.5 mmHg and raised splanchnic (+34%) and forearm (+70%) vascular resistances and HR (from 64 to 74 beats/min) without reducing MAP. NS plus LBNP reduced MAP from 103 +/- 8 to 95 +/- 6 mmHg and HR from 74 +/- 6 to 67 +/- 5 beats/min without changing CVP but did not alter vascular resistances, which remained elevated and constant throughout LBNP before and after NS. Increments in plasma concentrations of renin (240%), aldosterone (70%), epinephrine (112%), and norepinephrine (46%) accompanied LBNP and NS; a separate influence of NS was not discernible. We conclude that vasoconstriction in response to combined cardiopulmonary and aortic inhibition is not overpowered by carotid sinus stimulation.

Reflex control of the circulation during exercise.

Current theory is that circulatory control in exercise is governed by central command which sets basic patterns of effector activity that is modulated by arterial baroreflexes and chemo- and mechanoreflexes from active muscle. Because central command acts on vagal activity rather than sympathetic nerve activity (SNA), and because muscle chemoreflexes are not normally active during mild to moderate dynamic exercise, current theory cannot explain why SNA to virtually all organs, including active muscle, increases even during mild exercise. Are arterial baroreflexes involved? Baroreflex sensitivity is maintained during exercise, and most importantly, the reflex is reset to higher blood pressure (BP). A new hypothesis is that central command works by resetting the baroreflex to a higher BP and withdraws vagal activity to raise heart rate, cardiac output and BP at the onset of exercise. The key to the hypothesis is that the rise in cardiac output at exercise onset must be fast enough to raise BP to its new reset level immediately, otherwise a BP error occurs that must be corrected by baroreflex and SNA.

Cardiovascular responses to graded reductions in leg perfusion in exercising humans.

Our objective was to determine whether the chemoreflex from human muscle is elicited by small graded reductions in muscle blood flow (MBF) during mild exercise or whether this reflex has an obvious threshold associated with large changes in femoral venous lactate and H+ levels (i.e., as in dogs with high muscle oxidative capacity). Seven subjects exercised supine at 40, 87, and 142 W; lower body positive pressure (LBPP) was applied in 3-min steps at 25, 35, 45, and 50-60 mmHg with the lower body and the cycle ergometer in a sealed box. Estimated MBF (Fick) fell by 5.3 +/- 4.3 to 19.9 +/- 3.8% at four levels of LBPP over three work rates. Mean arterial pressure (MAP), heart rate (HR), and plasma norepinephrine (NE) concentration rose with increasing LBPP. MAP was significantly correlated with femoral venous pH, lactate, O2 tension, and O2 content during moderate and heavy exercise, without an apparent threshold. Percentage decreases in muscle vascular conductance exceeded the decreases in MBF twofold, indicating significant opposition to reduction in MBF by the chemoreflex. Approximately 50% of the correction of MBF back toward control (i.e., at 0 LBPP) could be explained by increased cardiac output, calculated from the rise in HR; the remaining correction could be attributed to both sympathetic vasoconstriction (indicated by high NE levels) and to mechanical effects of partial occlusion. Results suggest that in humans stepwise reductions in MBF gradually elicit muscle chemoreflexes with no apparent threshold at these levels of exercise.

Blood pressure regulation during exercise.

This brief review examines five problems concerning arterial blood pressure regulation during exercise. These are: 1. A history and summary of evidence that baroreflexes are, or are not, active during exercise. 2. What might be other "regulators" of blood pressure during exercise? The characteristics of a blood pressure-raising reflex from ischemic and active skeletal muscle (muscle chemoreflex) is reviewed along with a putative role for centrally generated motor command signals (central command). 3. How blood pressure is maintained during exercise. The importance of regional vasoconstriction, particularly in active skeletal muscle, is reviewed. 4. How well matched are cardiac output and total vascular conductance? Does demand for muscle blood flow outstrip cardiac pumping capacity? 5. Reflex control of blood pressure by both baroreflexes and muscle chemoreflexes. The importance of baroreflexes and evidence for resetting is reviewed. A new hypothesis is stated.

Baroreflex-induced vasoconstriction in active skeletal muscle of conscious dogs.

We investigated the magnitude of baroreflex-mediated vasoconstriction in the hindlimbs of six conscious dogs at rest and during four levels of treadmill exercise ranging in intensity from mild (2 mph, 0% grade) to heavy (6 mph, 10% grade). Dogs were instrumented with vascular occluders on both common carotid arteries, an electromagnetic flow probe and vascular occluder on the terminal aorta, and a catheter in a branch of the femoral artery; aortic baroreceptors were intact. The responses to a 2-min carotid occlusion were observed at rest and after 3-5 min of exercise at each work rate. The increases in mean arterial pressure during carotid occlusion were similar at rest and at each level of exercise (26 +/- 4 to 35 +/- 3 mmHg; no significant difference). At rest, carotid occlusion caused only a small but significant decrease in terminal aortic vascular conductance (TAC) (-0.89 +/- 0.21 ml.min-1.mmHg-1, P less than 0.05). During mild exercise, baseline terminal aortic blood flow (TAQ) and TAC increased, and the reduction in TAC during carotid occlusion exceeded that observed at rest (-1.85 +/- 0.42 ml.min-1.mmHg-1, P less than 0.05). As exercise intensity increased, the magnitude of the reduction in TAC during carotid occlusion increased linearly with the baseline TAQ. At the highest work rate, approximately 59% of the increase in mean arterial pressure during carotid occlusion was due to the large decrease in TAC (-6.35 +/- 0.50 ml.min-1.mmHg-1). We conclude that the vasoconstriction of active skeletal muscle during the pressor response to bilateral carotid occlusion increased with exercise intensity.(ABSTRACT TRUNCATED AT 250 WORDS)

Sympathetic activity during graded central hypovolemia in hypoxemic humans.

To determine how hypoxemia (Hx) might alter muscle sympathetic nerve activity (MSNA) (microneurography, peroneal nerve), norepinephrine (NE) levels, and vasoconstriction during mild central hypovolemia, we exposed eight men to continuous graded lower body negative pressure (LBNP) (-5, -10, -15, -20, and -25 mmHg, 5 min per level) during both Hx (10 or 12% O2) and normoxia (Nx). Hx significantly augmented MSNA during LBNP. Total MSNA (average amplitude X burst frequency) rose at each level of LBNP by 2, 28, 93, 61, and 123% (Nx) and 32, 110, 127, 179, and 216% (Hx). Only at LBNP -20 and -25 mmHg did Hx significantly augment the increase in forearm venous NE concentration. Arterial pressure was unaffected by LBNP in Nx and Hx. Forearm blood flow (venous occlusion plethysmography) fell, and forearm vascular resistance (FVR) rose 23, 53, 65, 67, and 86% (Nx) vs. 22, 23, 60, 69, and 87% (Hx), but increments in FVR (absolute units) were significantly less in Hx. Correlations among MSNA and other variables were insignificant for pooled data owing to large inter-individual variations in slopes, but correlations were significant for total MSNA (and burst frequency) vs. FVR (Nx) and NE (Hx). Three men released epinephrine during LBNP; this was accompanied by forearm vasodilation and falling pressure, and in two men, decreased MSNA and bradycardia occurred (i.e., vasovagal reaction). Overall, we found no major defect in sympathetic control during graded hypovolemia and Hx as long as epinephrine levels did not rise.

Reflex control of the circulation during exercise: chemoreflexes and mechanoreflexes.

The overall scheme for control is as follows: central command sets basic patterns of cardiovascular effector activity, which is modulated via muscle chemo- and mechanoreflexes and arterial mechanoreflexes (baroreflexes) as appropriate error signals develop. A key question is whether the primary error corrected is a mismatch between blood flow and metabolism (a flow error that accumulates muscle metabolites that activate group III and IV chemosensitive muscle afferents) or a mismatch between cardiac output (CO) and vascular conductance [a blood pressure (BP) error] that activates the arterial baroreflex and raises BP. Reduction in muscle blood flow to a threshold for the muscle chemoreflex raises muscle metabolite concentration and reflexly raises BP by activating chemosensitive muscle afferents. In isometric exercise, sympathetic nervous activity (SNA) is increased mainly by muscle chemoreflex whereas central command raises heart rate (HR) and CO by vagal withdrawal. Cardiovascular control changes for dynamic exercise with large muscles. At exercise onset, central command increases HR by vagal withdrawal and "resets" the baroreflex to a higher BP. As long as vagal withdrawal can raise HR and CO rapidly so that BP rises quickly to its higher operating point, there is no mismatch between CO and vascular conductance (no BP error) and SNA does not increase. Increased SNA occurs at whatever HR (depending on species) exceeds the range of vagal withdrawal; the additional sympathetically mediated rise in CO needed to raise BP to its new operating point is slower and leads to a BP error. Sympathetic vasoconstriction is needed to complete the rise in BP. The baroreflex is essential for BP elevation at onset of exercise and for BP stabilization during mild exercise (subthreshold for chemoreflex), and it can oppose or magnify the chemoreflex when it is activated at higher work rates. Ultimately, when vascular conductance exceeds cardiac pumping capacity in the most severe exercise both chemoreflex and baroreflex must maintain BP by vasoconstricting active muscle.

Cardiovascular responses to carotid sinus baroreceptor stimulation during moderate to severe exercise in man.

Our objective was to assess the importance of arterial baroreflexes in maintaining vasoconstriction in active muscle during moderate to severe exercise. Eight subjects exercised for 8-15 min on a cycle ergometer at three levels (averages 94, 194, 261 W) requiring 40-88% of VO2 max. Four times during each exercise level pulsatile negative pressure (-50 mmHg) was applied over the carotid sinuses for 30 s; suction was applied at each ECG R-wave for 250-400 ms. Before and during each neck suction, femoral venous blood flow (FVBF) was measured by constant infusion thermal dilution. At 94 W neck suction significantly reduced blood pressure (BP) (15 mmHg) and heart rate (HR) (7 beats min-1), and raised leg vascular conductance (LVC) (11.4%) without changing FVBF. At 194 W, neck suction reduced BP (9 mmHg), HR (4 beats min-1) and FVBF (5.1%, 240 ml min-1), and raised LVC (5.2%). At 261 W, LVC was unchanged by neck suction, but BP and FVBF both fell (9 mmHg and 650 ml min-1 or 7.4%). We conclude that competing local vasodilation and sympathetic vasoconstriction control muscle blood flow during moderate exercise, and vasoconstrictor tone can be withdrawn by baroreceptor stimulation. High levels of vasoconstrictor outflow to muscle in severe exercise may not originate from baroreflexes.

Baroreflex attenuates pressor response to graded muscle ischemia in exercising dogs.

Graded reductions in hindlimb perfusion in dogs exercising at 2 miles/h (0% grade) elicited reflex pressor responses by what is referred to as the "muscle chemoreflex." To determine the extent to which arterial baroreceptor reflexes oppose the muscle chemoreflex, we elicited pressor responses to muscle ischemia before and after chronic surgical denervation of the arterial baroreceptors. The muscle chemoreflex showed a threshold beyond which systemic pressure rose approximately 3 mmHg for each 1-mmHg decrease in hindlimb perfusion pressure when the arterial baroreceptors were intact. Arterial baroreceptor denervation approximately doubled the pressor responses, i.e., systemic pressure rose by approximately 6 mmHg for each 1-mmHg fall in hindlimb perfusion pressure, without alteration in threshold. We conclude that during mild dynamic exercise, the arterial baroreflexes oppose the pressor response to graded reductions in hindlimb perfusion, reducing it by approximately 50%. When unopposed by the arterial baroreflexes the muscle chemoreflex exhibits a gain (ratio of change in systemic pressure to change in hindlimb perfusion pressure) of approximately -6; thus this reflex can correct by 85% the decrease in muscle perfusion pressure caused by partial vascular occlusion.

Hypotension induced by central hypovolaemia and hypoxaemia.

Our question was whether the reduced orthostatic tolerance that accompanies hypoxaemia in some (not all) subjects might be associated with an abnormally large release of adrenaline. Eight normal young men were exposed to lower body negative pressure (LBNP) at -30 to -40 mmHg while breathing air or 10% O2 in N2. Four subjects developed hypotension and bradycardia whenever LBNP was applied during hypoxaemia; four showed a rise in heart rate and stable blood pressure. During normoxia plasma adrenaline concentration did not rise during LBNP in any subject, nor during hypoxaemia plus LBNP in the subjects who remained normotensive. In the four men whose heart rates and blood pressures fell during LBNP with hypoxaemia, adrenaline rose markedly, reaching 200-1600 pg ml-1. All subjects showed similar elevations in noradrenaline concentration during LBNP in both normoxia and hypoxaemia. The results suggest that reduced tolerance to central hypovolaemia during hypoxaemia could stem from known vasomotor and cardiac effects attending high plasma concentrations of adrenaline.

Hypoxemia raises muscle sympathetic activity but not norepinephrine in resting humans.

The experimental objective was to determine whether moderate to severe hypoxemia increases skeletal muscle sympathetic nervous activity (MSNA) in resting humans without increasing venous plasma concentrations of norepinephrine (NE) and epinephrine (E). In nine healthy subjects (20-34 yr), we measured MSNA (peroneal nerve), venous plasma levels of NE and E, arterial blood pressure, heart rate, and end-tidal O2 and CO2 before (control) and during breathing of 1) 12% O2 for 20 min, 2) 10% O2 for 20 min, and 3) 8% O2 for 10 min--in random order. MSNA increased above control in five, six, and all nine subjects during 12, 10, and 8% O2, respectively (P less than 0.01), but only after delays of 12 (12% O2) and 4 min (8 and 10% O2). MSNA (total activity) rose 83 +/- 20, 260 +/- 146, and 298 +/- 109% (SE) above control by the final minute of breathing 12, 10, and 8% O2, respectively. NE did not rise above control at any level of hypoxemia; E rose slightly (P less than 0.05) at one time only with both 10 and 8% O2. Individual changes in MSNA during hypoxemia were unrelated to elevations in heart rate or decrements in blood pressure and end-tidal CO2--neither of which always fell. We conclude that in contrast to some other sympathoexcitatory stimuli such as exercise or cold stress, moderate to severe hypoxemia increases leg MSNA without raising plasma NE in resting humans.

Does acute hypoxemia blunt sympathetic activity in hyperthermia?

Sympathetic alpha-adrenergic function is depressed by hypoxemia per se; does addition of another sympathoexcitatory stimulus elicit normal responses in other sympathetic effector pathways? We activated by hyperthermia four sympathetic pathways: alpha-adrenergic [norepinephrine (NE) release], beta-adrenergic [plasma renin activity (PRA)], cholinergic (sweating), and peptidergic (active vasodilation). In the first test, five normothermic men were exposed to hypoxemia for 10 min (control), then hypoxemia plus heat for 30 min, and then heat with normoxia for 8-10 min over a continuous 48- to 50-min period. Heating was controlled with a water-perfused suit. Time courses and magnitudes of heat-induced increments in body temperature, forearm blood flow, and sweat rate were normal during hypoxemia and unaffected by switching to normoxia. Hypoxemia exaggerated increases in plasma NE, epinephrine, PRA, and heart rate but had no additional effects on blood pressure. In a second 50-min test (2 men) with normoxic control (10 min), heating plus normoxia (20 min), and heating plus hypoxemia (20 min), effects of hypoxemia on all variables were as in the first test. Thus, acute moderate hypoxemia did not blunt active cutaneous vasodilation or sweating and exaggerated increases in catecholamines and heart rate, indicating maintained peripheral autonomic function.

Venomotor responses during central and local hypoxia.

Venomotor responses, measured as the pressure rise in occluded forearm veins, were used in a two-part experiment to test presence or absence of sympathetic neuroeffector mechanisms in 10 men made moderately to severely hypoxemic. In part I, forearm venoconstriction was induced by ice water on the contralateral forearm (a spinal reflex) in eight supine, resting men who breathed air, 10.3% oxygen or 7.7% oxygen. Large reflex venoconstrictions persisted during hypoxia. In part II (seven men), venoconstriction was centrally induced by exercise while subjects were 1) normoxic; 2) arm hypoxic, body normoxic; 3) arm hyperoxic (or normoxic), body hypoxic; or 4) both arm and body hypoxic. Arm vs. body oxygen tensions were separated by occluding the arm as one gas mixture was breathed, then switching the subject to another mixture as the arm remained occluded. Strong venoconstrictor responses to moderate exercise (100-150 W) persisted during both local and central hypoxemia. We conclude that moderate to severe hypoxemia does not block, pre- or postjunctionally, sympathetic venoconstriction that originates from spinal reflexes (cold). Venoconstriction in exercise (presumably originating in higher centers) was not blocked by moderate hypoxemia; severe hypoxemia was not studied.