doi:10.1152/ajpregu.00662.2010 300:R663-R673, 2011. First published 22 December 2010;Am J Physiol Regul Integr Comp Physiol

Barker and José González-AlonsoJames Pearson, David A. Low, Eric Stöhr, Kameljit Kalsi, Leena Ali, Horacetemperature on skeletal muscle blood flowand exercising human leg: insight into the effect of Hemodynamic responses to heat stress in the resting

You might find this additional info useful...

46 articles, 36 of which can be accessed free at:This article cites http://ajpregu.physiology.org/content/300/3/R663.full.html#ref-list-1

including high resolution figures, can be found at:Updated information and services http://ajpregu.physiology.org/content/300/3/R663.full.html

can be found at:and Comparative PhysiologyAmerican Journal of Physiology - Regulatory, Integrativeabout Additional material and information

http://www.the-aps.org/publications/ajpregu

This infomation is current as of May 16, 2011. 

ISSN: 0363-6119, ESSN: 1522-1490. Visit our website at http://www.the-aps.org/.Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2011 by the American Physiological Society. ranging from molecules to humans, including clinical investigations. It is published 12 times a year (monthly) by the Americanilluminate normal or abnormal regulation and integration of physiological mechanisms at all levels of biological organization,

publishes original investigations thatAmerican Journal of Physiology - Regulatory, Integrative and Comparative Physiology

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

Hemodynamic responses to heat stress in the resting and exercising humanleg: insight into the effect of temperature on skeletal muscle blood flow

James Pearson,1 David A. Low,1 Eric Stöhr,1 Kameljit Kalsi,1 Leena Ali,2 Horace Barker,2

and José González-Alonso1

1Centre for Sports Medicine and Human Performance, Brunel University West London, Uxbridge; and 2Departmentof Anaesthetics, Ealing Hospital National Health Service Trust, Southall, Middlesex, United Kingdom

Submitted 5 October 2010; accepted in final form 21 December 2010

Pearson J, Low DA, Stöhr E, Kalsi K, Ali L, Barker H,González-Alonso J. Hemodynamic responses to heat stress in theresting and exercising human leg: insight into the effect of tempera-ture on skeletal muscle blood flow. Am J Physiol Regul Integr CompPhysiol 300: R663–R673, 2011. First published December 22, 2010;doi:10.1152/ajpregu.00662.2010.—Heat stress increases limb bloodflow and cardiac output (Q) in humans, presumably in sole responseto an augmented thermoregulatory demand of the skin circulation.Here we tested the hypothesis that local hyperthermia also increasesskeletal muscle blood flow at rest and during exercise. Hemodynam-ics, blood and tissue oxygenation, and muscle, skin, and core temper-atures were measured at rest and during exercise in 11 males acrossfour conditions of progressive whole body heat stress and at restduring isolated leg heat stress. During whole body heat stress, legblood flow (LBF), Q, and leg (LVC) and systemic vascular conduc-tance increased gradually with elevations in muscle temperature bothat rest and during exercise (r2 � 0.86–0.99; P � 0.05). EnhancedLBF and LVC were accompanied by reductions in leg arteriovenousoxygen (a-vO2) difference and increases in deep femoral venous O2

content and quadriceps tissue oxygenation, reflecting elevations inmuscle and skin perfusion. The increase in LVC occurred despite anaugmented plasma norepinephrine (P � 0.05) and was associated withelevations in muscle temperature (r2 � 0.85; P � 0.001) and arterialplasma ATP (r2 � 0.87; P � 0.001). Isolated leg heat stress accountedfor one-half of the increase in LBF with severe whole body heat stress.Our findings suggest that local hyperthermia also induces vasodilata-tion of the skeletal muscle microvasculature, thereby contributing toheat stress and exercise hyperemia. The increased limb muscle vaso-dilatation in these conditions of elevated muscle sympathetic vaso-constrictor activity is closely related to the rise in arterial plasma ATPand local tissue temperature.

vasodilatation; hyperthermia; ATP

HEAT STRESS AUGMENTS LIMB blood flow and cardiac output (Q)in resting humans (1, 4, 9, 19, 35, 40, 45). An unresolvedquestion is whether limb muscle vasodilatation contributes tothese increases in blood flow and the mechanisms involved.Early investigations into the partition of limb perfusion be-tween skin and skeletal muscle during heat stress producedconflicting results (4–6, 9, 35), with some studies suggestingan elevation in muscle blood flow (4, 5). However, later studiesutilizing a variety of experimental techniques found no evi-dence of an elevation in forearm muscle perfusion during heatstress (2, 8, 19). These observations combined with the esti-mate of maximal skin blood flow (SkBF) of 6–8 l/min, basedon indirect measures of Q and visceral blood flow during whole

body heat stress, helped shape the view that heat stress-inducedsystemic hyperemia and blood flow redistribution are entirelydue to an augmented thermoregulatory demand for SkBF (8,25, 39, 40). Recent evidence, however, shows that calf muscleblood flow increases significantly during isolated leg heating(25). Nevertheless, the mechanisms increasing skeletal muscleperfusion with passive heat stress and the contribution of localvs. systemic elevations in tissue temperature on limb muscleblood flow regulation remain unknown.

The effects of combined heat stress and exercise upon activemuscle blood flow also remain equivocal with reports showingthat exercising limb and systemic perfusion is either elevated(42, 47), unchanged, or reduced (10, 28, 29, 41). These dis-crepant findings may be accounted for by differences in themode and intensity of exercise, the magnitude of heat stress,and the possible influence of dehydration. In this context, limbperfusion is elevated during single limb exercise with exposureto a moderate degree of either local limb or whole body heatstress in conditions where dehydration is negligible (42, 47).On the other end of the spectrum, limb muscle perfusion isreduced in association with the hemoconcentration and thedeclines in perfusion pressure and Q that accompany severeheat stress or dehydration during prolonged and short moder-ate- to high-intensity whole body exercise (10, 11). The use ofan isolated limb exercise model and the maintenance of thesubject’s hydration status enable the study of the influence ofgraded levels of heat stress on exercising limb muscle bloodflow without the confounding influences of reflexes underpin-ning the circulatory limitations to whole body exercise anddehydration (3, 26).

Mechanistically, the elevation in limb perfusion with heatstress occurs in the presence of increased muscle sympatheticnerve activity (30, 33). This indicates that signals responsive toheat stress may directly or indirectly modulate muscle sympa-thetic vasoconstrictor activity such that vasodilatation prevailsover vasoconstriction, allowing limb muscle perfusion andvascular conductance to increase (21). The observation that theplasma concentration of the potent vasodilator and sympatho-lytic molecule ATP increases during exercise with severe heatstress (12, 36) raises the possibility that ATP is one of the keyregulatory signals involved in the prevailing heat stress-in-duced limb vasodilatation. It is yet to be determined whetherincreases in local temperature induce elevations in limb muscleperfusion in relation to the rise in circulating ATP.

Accordingly, the main aim of this study was to determinewhether skeletal muscle vasodilatation contributes to the gen-erally observed increases in limb and systemic perfusion withheat stress in resting humans and whether heat stress inducessimilar effects on leg muscle and systemic perfusion during

Address for reprint requests and other correspondence: J. González-Alonso,Centre for Sports Medicine and Human Performance, Brunel Univ., Uxbridge,Middlesex UB8 3PH, UK (e-mail: j.gonzalez-alonso@brunel.ac.uk).

Am J Physiol Regul Integr Comp Physiol 300: R663–R673, 2011.First published December 22, 2010; doi:10.1152/ajpregu.00662.2010.

0363-6119/11 Copyright © 2011 the American Physiological Societyhttp://www.ajpregu.org R663

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

mild exercise in euhydrated individuals. A second aim was todetermine the contribution of local tissue temperatures to theincreases in leg perfusion with whole body and isolated legheat stress and gain insight into the role of plasma ATP andmuscle temperature into heat stress-mediated limb musclevasodilatation. To accomplish these aims, we measured leg andsystemic hemodynamics, arterial, deep, and common femoralvenous blood O2 content, quadriceps muscle oxygenation andtemperature, and plasma ATP and catecholamines at rest andduring moderate one-legged, knee-extensor exercise. Thesemeasurements were performed in healthy male volunteersduring control conditions and three graded levels of wholebody skin and internal body hyperthermia as well as duringisolated leg hyperthermia. The design of the study allowed usto test the overall hypothesis that local hyperthermia inducesvasodilatation in resting and exercising human limb muscleand gain an insight into the roles of muscle temperature andintravascular ATP in this response.

METHODS

Subjects. Eleven healthy, recreationally active males (mean � SD:ages, 21 � 2 yr; body wt, 76.3 � 10.4 kg; and height, 178 � 6 cm)participated in this study involving two different protocols. This studyconformed to the code of Ethics of the World Medical Association(Declaration of Helsinki) and was conducted after ethical approvalfrom the Brunel University Research Ethics Committee. Informedwritten and verbal consent was obtained from all participants beforeparticipation.

Experimental protocols. In protocol 1 leg and systemic hemody-namics and quadriceps muscle oxygenation were examined at rest andduring moderate one-legged, knee-extensor exercise (mean � SE:21 � 1 W at 65 � 1 rpm for 6 min) in four consecutive whole body,heat-stress conditions. To induce whole body heat stress, a custom-built, water-perfused suit, passively heated participants in the supineposition to four progressive levels of heat stress leading to fourdistinct conditions of skin and internal body hyperthermia: 1) control:normal skin, muscle, and core temperatures (i.e., �33, �34, and�37°C, respectively); 2) Skin Hyperthermia: whole body skin andmuscle temperatures were increased, while rectal temperature re-mained at control (all �37°C); 3) Skin and Mild Core Hyperthermia:whole body skin, muscle, and rectal temperatures increased (all�38°C); and 4) Skin and Core Hyperthermia: skin temperatureremained elevated and core and muscle temperatures increased further(�38, �39, and �39°C, respectively) (Fig. 1). The same level of skinand core hyperthermia was maintained during exercise, yet quadricepsmuscle temperature increased in all four experimental conditions(�1–2°C) in response to the ensuing increases in local metabolic heatproduction and, in the control condition, the additional convectiveheat transfer from the warmer body core to the cooler leg muscles.The manipulation of body temperatures in this manner allowed theexamination of the contribution of muscle, skin, and core temperatureon leg hemodynamic responses to heat stress both at rest and duringexercise.

In protocol 2 leg and systemic hemodynamics were examinedusing noninvasive methods in the supine position at rest over 60 minof isolated leg heat stress, which created a condition of isolated legSkin Hyperthermia (i.e., core and experimental leg skin and muscletemperatures of �37°C). This protocol allowed the examination ofthe independent contribution of leg muscle and skin temperatures onthe leg hemodynamic response to whole body heat stress. Protocolswere separated by at least 2 wk. Subjects ingested a carbohydrate-electrolyte beverage (Gatorade) throughout both protocols to preventdehydration. The temperature of the beverage was 35–40°C to avoidany decreases in internal temperature caused by its consumption.

Instrumentation of subjects. In preparation for the whole body heatstress protocol, subjects were familiarized with the one-legged, knee-extensor exercise to minimize the involvement of the gluteal andhamstring muscles and thereby isolating the work load to the kneeextensors. Furthermore, all participants reported to the laboratory ontwo occasions separated by 2 days to undergo heat familiarization byperforming cycling exercise at �150 W for 60 min in a heat chambercontrolled at 37°C and 60% humidity.

On the morning of the protocol, subjects arrived at the laboratoryafter eating a light breakfast. After insertion of the rectal thermister,participants rested in the supine position, while catheters were placedunder local anesthesia into the common femoral vein of the exercisingleg (left leg) and in the radial artery (right forearm). The femoralvenous catheter was positioned 1–2 cm proximal to the inguinalligament, running proximally in the anterograde direction. In four outof the eleven participants, an additional catheter was inserted in thefemoral vein of the exercising leg that ran in a distal or retrogradedirection. This retrograde venous catheter was 12 cm in length andwas inserted into the femoral vein �1–2 cm from the anterogradecommon femoral venous catheter. Consequently the catheter tip wasinserted past the saphenofemoral junction, resting in the deep portionof the femoral vein. Consequently, the blood drawn from this catheterprimarily represented blood-draining leg skeletal muscle tissues andas such was not representative of blood flowing through the greatsaphenous vein. For the purposes of this publication, the retrogradefemoral venous catheter will be referred to as the deep femoral venouscatheter. Successful placement of the retrograde catheter in the deepfemoral venous portion of the leg was indicated by a lower blood O2

content compared with that drawn from the common femoral venouscatheter. Specifically, the absolute venous O2 content values in thedeep vs. mixed femoral venous catheter were �123 vs. �136 ml/l atrest and �62 vs. �70 ml/l during exercise, respectively. These valuesfrom the deep femoral catheter closely agree with previous reports(44) and can be used as an index of deep leg tissue oxygenation andperfusion.

Participants then walked to the experimental room and sat on theknee-extensor ergometer where they were dressed in a custom-built,water-perfused suit that was interwoven with silicone tubing andconnected to a water circulator (model F34; Julabo, Seelbach, Ger-many). The water circulator was fitted with an auxiliary pump andtemperature-control unit capable of controlling the temperature of thewater in the suit, which covered the subject’s entire body except thehead, hands, and feet. Whole body heat stress was induced byperfusing water at a temperature of 47°C through the water-perfusedsuit. Once the specific skin and/or core temperatures were attained ateach stage of heat stress, the temperature of the water-perfusing thesuit was decreased slightly to limit further increases in skin and/orcore temperature during data collection at rest and during exercise. Tominimize heat loss during each heat-stress stage, a thermal foil blanketcovered the torso and was wrapped around the lower body of thesubjects, socks covered both feet, and a thermal hat was also worn.After the participants were dressed in the suit they lay supine on areclining chair that was part of a knee-extensor ergometer (modelLE220; FBJ Engineering, Odense Area, Denmark), while the left footand ankle were inserted into the boot of the ergometer. Both of thesubject’s lower legs were supported during resting conditions.

In protocol 2, subjects only wore the left leg of the water-perfusedsuit. Similarly to protocol 1, foil was wrapped around the heated legand the water circulator controlled the temperature of the water withinthe water-perfused leg. Subjects remained supine while a pressure cuffwas placed around a finger for the measurement of systemic bloodpressure and subsequent determination of systemic hemodynamics.

Temperature measurements. Skin thermisters were placed on sevensites: forehead, forearm, hand, abdomen, thigh, calf, and foot (GrantInstruments, Cambridge, United Kingdom). Thermisters were se-curely held in place throughout the protocol by the use of adhesivespray and medical tape. Rectal temperature was measured 10 cm past

R664 LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

the sphincter muscle using a commercially available rectal probe(Physitemp, Clifton, NJ). Skin (Squirrel 1000 Series, Grant Instru-ments) and rectal (Thermalert; Physitemp) temperatures were moni-tored online. Weighted mean skin temperature was calculated usingmethods described previously (17). Mean leg skin temperature wascalculated using the combined weighted skin temperatures of the thighand calf.

In the whole body and isolated leg heat stress protocols, quadricepsmuscle temperature was measured online (model TC-2000; SableSystems, Las Vegas, NV) in five and two participants, respectively,with a tissue-implantable thermocouple microprobe (model T-204A;Physitemp) together with core and skin temperatures. Muscle temper-ature was measured in the vastus lateralis muscle at a depth of 3 cm.Increases in vastus lateralis muscle temperature provide an accuraterepresentation of the mean temperature responses of the differentportions of the quadriceps muscles during isolated knee-extensorexercise in conditions where subcutaneous tissue temperature remainsunchanged despite external heating (24).

Systemic hemodynamics and muscle oxygenation. In protocol 1,baseline systemic and leg hemodynamics were measured immediatelyprior to exercise after a minimum of 10-min supine rest and followingthe attainment of the desired skin and rectal temperatures. Duringexercise, these measurements were repeated between minutes 4 and 6.Additionally, arterial and venous blood samples (1 ml for blood gas,metabolite, and electrolyte variables; 2 ml for plasma ATP and plasmahemoglobin; and 2 ml for plasma catecholamines) were obtained atrest and after 5 min of exercise. Arterial and venous catheters werealso used to measure arterial and venous blood pressure, respectively.In protocol 2 (n � 7), subjects remained at rest throughout the isolatedleg heat stress protocol and temperature and hemodynamic measureswere taken every 2 min between 0–10 min and every 10 minthereafter.

In protocol 1, heart rate was obtained from a three-lead electrocar-diogram, while arterial and femoral venous pressure waveforms werecontinuously recorded at the level of the heart via pressure transducers(Pressure Monitoring Kit; Baxter) connected to two amplifiers (BPamp; ADInstruments, Bella Vista, NSW, Australia) and monitoredonline via a data acquisition system (Powerlab 16/30 ML 880/P;ADInstruments). Q was calculated as the product of heart rate andstroke volume, where stroke volume was estimated using directlymeasured arterial pressure waveforms via the Modelflow method,incorporating age, gender, height, and weight (BeatScope version 1.1;Finapress Medical Systems, Amsterdam, Netherlands) (46). Q dataobtained from estimated stroke volume via the Modelflow method inthis study, shares a tight relationship with an independently derivedmeasure of Q data obtained using echocardiography (r2 � 0.831; P �0.002; data not shown). Systemic vascular conductance was calcu-lated by dividing Q by mean arterial blood pressure (MAP).

Muscle oxygenation of the vastus lateralis muscle was measuredusing near-infrared spectroscopy (NIRS) (INVOS Cerebral Oximeter;Somanetics, Troy, MI). An adhesive NIRS pad that emitted NIRsignals at 730- and 810-nm wavelengths was placed over the vastuslateralis muscle. The NIRS pad contained two optodes at a distance of3 and 4 cm away from the sensor. Previous research (18) indicates thatthese NIRS signals penetrated quadriceps tissue to a depth of �1.5and 2 cm, respectively. Therefore, the shallow penetrating NIRSsignal enabled an algorithm inherent to the NIRS system to accountfor any alterations in cutaneous oxygenation. The NIRS pad was

Fig. 1. Body temperature responses to whole body heat stress. Mean wholebody skin temperature increased from control with the skin hyperthermiacondition, but core temperature remained unchanged. With further heat stress,core temperature increased in the skin and mild core hyperthermia condition,while mean whole body skin temperature remained elevated. This patterncontinued into the skin and core hyperthermia condition where there was afurther elevation in core temperature, and skin temperature remained elevatedfrom control. Data are means � SE for 11 subjects, except for muscletemperature (n � 5). Quadriceps muscle temperature followed the increase inmean skin temperature at rest and increased rapidly during exercise by 1–2°Cabove corresponding resting values. *Different from control, P � 0.05;#different from skin hyperthermia, P � 0.05; †different from skin and mildcore hyperthermia. Significance was accepted at P � 0.05 and refers todifferences in the respective conditions, i.e., either rest or exercise.

R665LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

securely taped to the skin to ensure that no light interfered withrecordings.

In protocol 2, blood pressure waveforms were recorded noninva-sively by using photosphygmomanometry (Finapres Medical Sys-tems, Smart Medical, Amsterdam, Netherlands) and heart rate wasobtained from a three-lead electrocardiogram, allowing estimates ofsystemic hemodynamics as described above. In both studies, systemicoxygen uptake was continuously measured and recorded online(Quark b2; Cosmed, Italy).

Leg and skin hemodynamics. In both studies, leg blood flow (LBF)was measured from the common femoral artery 2–3 cm proximal tothe bifurcation point by using an ultrasound Doppler equipped with a10-MHz linear probe (Vivid 7 Dimension; GE Medical, Horton,Norway). Femoral artery vessel diameter was determined after ob-taining three 2D images in the longitudinal view at �40 frames/s,depending on artery depth. The diameter was calculated using mea-surements obtained from the systolic and diastolic phases, whichaccounted for one-third and two-thirds of the vessel diameter, respec-tively (32). Mean blood velocity (Vmean) was calculated from aninsonation angle that was consistently � 60 degrees (32) and with thesample volume positioned in the center of the femoral artery. Bloodvelocity was calculated from the average of three measurements eachconsisting of between 10 and 12 velocity profiles. The contribution ofturbulence occurring at the vascular wall to blood flow measurementwas reduced by using a low-velocity rejection filter. Finally, LBF(l/min) was calculated using the equation: Vmean � � (vessel diameter/2)2 � 6 � 104.

Our coefficient of variation for measures of LBF during fourcontrol rest and exercise conditions is 8.1% at rest and 4.9% duringexercise (n � 5), which is small compared with the blood flowincreases evoked by heat stress. This control study also showed thatresting and exercising LBF is unchanged during four control rest andexercise conditions mimicking the experimental design of the presentstudy.

During whole body and isolated leg heat stress SkBF of theexercising leg was measured via laser-Doppler flowmetry (PerifluxFlowmetry System, Jarfalla, Sweden). The probe was secured to theskin of the thigh (i.e., above the vastus lateralis) and was not coveredby or in contact with the water-perfused suit.

Perfusion pressure at the level of the leg was calculated as MAPminus femoral venous pressure, where each pressure value wasobtained from the integration of the corresponding pressure curves.Leg vascular conductance (LVC) was calculated as LBF divided byperfusion pressure. Leg arteriovenous oxygen (a-vO2) difference wasthe difference in arterial and femoral venous blood O2 content, whileleg O2 delivery was the product of arterial O2 content and LBF. LegO2 extraction was the ratio between leg a-vO2 difference and arterialO2 content, while leg Q was calculated by multiplying LBF by lega-vO2 difference. Finally, leg tissue blood flow (i.e., LBF-saphenousBF) was calculated in each experimental condition using the Fickprinciple: leg tissue blood flow � leg tissue VO2/deep femoral-basedleg a-vO2 difference, where, leg tissue VO2 � leg VO2 � estimated legskin VO2, assuming that leg skin VO2, i.e., 1.5 ml/min; (1, 43)remained constant with graded heat stress and/or exercise.

Blood and plasma parameters. Blood gas variables, hemoglobinconcentration, metabolites, electrolytes, and osmolality were mea-sured using an automated analyzer (model ABL 825; Radiometer,Copenhagen, Denmark). Plasma ATP was determined with the lucife-rin-luciferase technique, using a luminometer with three automaticinjectors (Orion Microplate Luminometer; Berthold Detection Sys-tem, Pforzheim, Germany). Blood samples (2.0 ml) were drawn intoa stop solution (2.7 ml) containing S-(4-nitrobenzyl)-6-thioinosine (5nM), IBMX (100 �M), forskolin (10 �M), EDTA (4.15 mM), NaCl(118 mM), KCl (5 mM), and tricine buffer (40 mM) (16). Immediatelythereafter, the samples were centrifuged for 3 min at 4,000 g in plastictubes containing lithium heparin with an inert gel barrier for plasmaseparation (BD, Franklin Lakes, NJ) and measured in duplicates at

room temperature (20–22°C) using an ATP kit (ATP Kit SL; Bio-Therma, Dalarö, Sweden) with an internal ATP standard procedure.As an indicator of hemolysis, plasma hemoglobin was measuredspectrophotometrically (model 3500; Jenway, Essex, UK). Plasmalevels of the catecholamines epinephrine and norepinephrine weremeasured using the catecholamines assay kit Bi-CAT EIA (model17-EA613–192; Alpco Diagnostics) according to the manufacturer’sinstructions.

Statistics. A one-way repeated-measures ANOVA was performedon all dependent variables to test significance among the control andthree conditions of heat stress separately at rest and during exercise.When a significant difference (P � 0.05) was found, post hoc analysisof the data was conducted using Tukey’s honestly significant differ-ence test and applying a Bonferroni correction where appropriate(P � 0.0125). Where applicable, relationships were determined usingPearson’s product moment correlation using the mean data of allparticipants from each thermal stage (P � 0.05). Stepwise regressionanalysis was also used to identify the best predictors of the LVCresponses to heat stress and exercise.

RESULTS

Hydration and temperature during whole body heat stress.In protocol 1 body weight, blood electrolytes, osmolality, andhematological variables remained unchanged (Table 1), indi-cating a maintained intravascular and extravascular fluid statusduring heat stress. With Skin Hyperthermia, mean skin tem-perature increased from 32.6 � 0.2°C to 36.8 � 0.2°C (P �0.05), whereas core temperature was unchanged (37.1 � 0.1°Cvs. 37.1 � 0.1°C; P 0.05). In the Skin and Mild CoreHyperthermia condition, mean skin temperature was main-tained (37.7 � 0.3°C), while core temperature increased to38.0 � 0.1°C (P � 0.05). This pattern was repeated withfurther whole body heat stress in the Skin and Core Hyper-thermia condition: 38.3 � 0.3°C (mean skin) and 38.6 � 0.1°C(core) (both P � 0.05). In response to whole body heat stress,mean leg skin temperature, and quadriceps muscle temperaturefollowed the same pattern and magnitude as mean whole bodyskin temperature (Fig. 1). In five participants, quadricepsmuscle temperature increased progressively from 35.3 � 0.4°Cat control rest to 37.0 � 0.1°C with Skin Hyperthermia, 37.9 �0.3°C with Skin and Mild Core Hyperthermia, and finally38.8 � 0.4°C with Skin and Core Hyperthermia (all P � 0.05).During exercise, muscle temperature increased (P � 0.05) andwas progressively higher with each thermal condition; from37.3 � 0.3°C at control, to 38.1 � 0.2°C, 38.9 � 0.3°C, and39.6 � 0.3°C, respectively (all P � 0.05). With the exceptionof quadriceps muscle temperature, all reported resting temper-atures are representative of exercise conditions as skin or coretemperatures were not significantly different between rest andexercise (P 0.05) (Fig. 1).

Leg and systemic hemodynamics and oxygenation duringwhole body heat stress. At rest, LBF and SkBF graduallyincreased with each level of whole body hyperthermia accom-panying a decline in leg a-vO2 difference and a significant butsmall increase in leg and whole body VO2 (peak VO2 �0.008 � 0.005 and 0.15 � 0.03 l/min; respectively, P � 0.05,Fig. 2). The increase in LBF was due to increases in meanblood velocity as common femoral artery diameter remainedunchanged throughout heat stress (overall mean: 0.93 � 0.02cm; P � 0.15). Increases in LBF were accompanied byelevations in common femoral venous O2 content and PO2

(136 � 7 vs. 164 � 2 ml/l and 38 � 2 vs. 50 � 2 mmHg,

R666 LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

respectively; P � 0.05, Fig. 3). In four participants, theseincreases in LBF and common femoral venous O2 content wereaccompanied by elevations in deep femoral venous O2 contentand PO2 (123 � 15 vs. 157 � 4 ml/l, P � 0.05, and 36 � 4 vs.43 � 5 mmHg, respectively, Fig. 3).

Perfusion pressure declined during whole body heat stressowing to a fall in MAP (P � 0.05), while femoral venouspressure slightly increased (11 � 1 vs. 14 � 3 mmHg; P �0.05) (Table. 2). At the level of the systemic circulation,both Q and systemic vascular conductance increased pro-gressively with each level of whole body hyperthermia(Table 2) accompanying gradual increases in heart rate (P �0.05).

During exercise, LBF increased with Skin Hyperthermiacompared with control (2.27 � 0.16 vs. 1.89 � 0.15 l/min P �0.05). Thereafter, LBF remained elevated from control withSkin and Mild Core Hyperthermia (2.48 � 0.25 l/min P � 0.05vs. control) and increased further with Skin and Core Hyper-thermia (2.55 � 0.23 l/min P � 0.05 vs. control and Skin andMild Core Hyperthermia, Fig. 2). The increase in LBF wasaccompanied by proportional decreases in leg a-vO2 difference(P � 0.05). In line with this, LVC increased from 14 � 1 to18 � 1 and 22 � 2 ml·min�1·mmHg�1 from control to SkinHyperthermia and Skin and Mild Core Hyperthermia condi-tions, respectively (P � 0.05), while no further increase wasobserved with Skin and Core Hyperthermia. Likewise, Q, heart

rate, and systemic vascular conductance progressively in-creased during exercise with heat stress (P � 0.05, Fig. 4).MAP declined from control values of 146 � 2 to 129 � 2mmHg with Skin Hyperthermia and was further reduced to120 � 3 mmHg with Skin and Mild Core Hyperthermia (P �0.05; Table 2). Thereafter, MAP remained stable as did leg andwhole body Q. Additionally, while vastus lateralis oxygenationdecreased with the onset of exercise with control and SkinHyperthermia conditions, leg SkBF increased rapidly (P �0.05; Table 2).

Effect of whole body heat stress at rest vs. exercise. LBFincreased in line with muscle temperature at rest (r2 � 0.99;P � 0.05; 0.34 � 0.02 l·min�1·°C�1) and to a similar magni-tude during exercise (r2 � 0.99; P � 0.05; 0.47 � 0.04l·min�1·°C�1). Similarly, increases in Q shared a strong rela-tionship with changes in muscle temperature at rest and duringexercise (r2 � 0.93; P � 0.05; 1.09 � 0.21 l·min�1·°C�1 andr2 � 0.99; P � 0.05; 1.40 � 0.08 l·min�1·°C�1, respectively).Furthermore, at rest and during exercise, elevations in muscletemperature were accompanied by increases in LVC (r2 �0.86, P � 0.001). However, when comparing the effect ofwhole body heat stress at rest vs. during exercise, the increasesin LBF and Q (and muscle temperature) with severe heat stresscompared with control were significantly attenuated duringexercise compared with that at rest (i.e., LBF � 0.66 � 0.15vs. 1.10 � 0.10 l/min and Q � 3.4 � 0.3 vs. 4.0 � 0.3 l/min,

Table 1. Blood variable responses to whole body heat stress at rest and during exercise

Control Skin Hyperthermia Skin and Mild Core Hyperthermia Skin and Core Hyperthermia

Rest Exercise Rest Exercise Rest Exercise Rest Exercise

Hb, g/l a 146 � 3 152 � 3 145 � 3 150 � 3 147 � 3 151 � 3 149 � 3 152 � 3v 147 � 3 156 � 3 148 � 3 152 � 3 150 � 3 153 � 3 151 � 3 153 � 3

O2 Sat, % a 97.9 � 0.2 97.9 � 0.3 97.8 � 0.2 98.1 � 0.1 98.0 � 0.3 98.0 � 0.2 98.6 � 0.2 98.3 � 0.2v 67.1 � 3.5 32.3 � 1.5 80.5 � 1.5 37.0 � 1.5 83.5 � 1.6 41.8 � 2 83.7 � 1.7 46.1 � 1.9

PO2, mmHg a 109 � 4 107 � 4 102 � 3 110 � 4 110 � 5 110 � 5 121 � 5†# 113 � 5v 38 � 2 25 � 1 48 � 2 26 � 1 50 � 2* 28 � 1* 48 � 2* 29 � 1*

CtO2, ml/l a 199 � 5 209 � 5 197 � 5 204 � 5 200 � 7 206 � 5 204 � 6# 207 � 6v 128 � 8 65 � 4 165 � 3* 75 � 4 170 � 5* 87 � 6* 174 � 7* 99 � 5*

PCO2, mmHg a 39 � 2 39 � 2 39 � 1 35 � 1 32 � 2* 31 � 3 25 � 3* 29 � 3v 49 � 1 67 � 2 42 � 1* 58 � 3 36 � 2*# 50 � 3 29 � 3*†# 43 � 4

pH a 7.41 � 0.02 7.40 � 0.02 7.40 � 0.01 7.43 � 0.02 7.44 � 0.02 7.45 � 0.02 7.52 � 0.03*# 7.48 � 0.03v 7.37 � 0.01 7.27 � 0.01 7.40 � 0.01* 7.31 � 0.02* 7.43 � 0.01*# 7.37 � 0.02*# 7.50 � 0.02*†# 7.40 � 0.03

Na�, mmol/l a 139 � 1 142 � 2 139 � 1 142 � 3 138 � 2 140 � 2 137 � 2 139 � 2v 139 � 2 143 � 1 139 � 2 142 � 3 138 � 1 141 � 0 137 � 2 139 � 2

K�, mmol/l a 4.0 � 0.1 4.6 � 0.1 4.0 � 0.1 4.4 � 0.1 3.9 � 0.1 4.1 � 0.1 3.9 � 0.1 4.0 � 0.1v 4.1 � 0.1 4.9 � 0.1 4.0 � 0.1 4.7 � 0.1 3.9 � 0.1 4.3 � 0.1 3.8 � 0.1 4.1 � 0.1

Cl�, mmol/l a 108 � 3 108 � 1 108 � 1 108 � 1 107 � 1 108 � 1 107 � 1 105 � 1v 105 � 2 105 � 1 106 � 1 104 � 1 106 � 1 104 � 1 105 � 1 104 � 1

Glucose, mmol/l a 5.9 � 0.2 5.8 � 0.1 7.5 � 0.3 7.2 � 0.3 9.2 � 0.5 8.2 � 0.4 8.7 � 0.5 7.9 � 0.4v 5.5 � 0.2 5.7 � 0.1 7.0 � 0.3 7.0 � 0.3 8.5 � 0.4 7.8 � 0.4 8.1 � 0.5 7.5 � 0.4

Lactate, mmol/l a 1.0 � 0.1 2.7 � 0.3 1.6 � 0.2 2.6 � 0.2 2.9 � 0.3 3.3 � 0.3 3.0 � 0.3 3.7 � 0.5v 1.0 � 0.1 4.1 � 0.5 1.5 � 0.1 3.4 � 0.5 2.8 � 0.3 3.8 � 0.5 3.0 � 0.3 4.1 � 0.7

Osmolality , mOsm·kg�1 a 284 � 2 289 � 3 283 � 2 288 � 3 285 � 2 286 � 2 282 � 3 285 � 3v 284 � 2 291 � 2 285 � 3 291 � 3 284 � 2 289 � 3 281 � 2 285 � 3

ATP, nmol/l a 454 � 102 709 � 158§ 639 � 120*§ 842 � 193§ 704 � 150§ 950 � 227§ 917 � 196§ 1090 � 200*†§v 685 � 80 867 � 114 754 � 42 851 � 78 824 � 115 903 � 55 874 � 155 822 � 145

Noradrenaline, nmol/l a 0.7 � 0.3 1.3 � 0.4 0.6 � 0.3 1.1 � 0.3 1.5 � 0.6 2.6 � 0.5 2.0 � 0.7* 4.6 � 1.4*v 0.8 � 0.2 2.1 � 1.3 0.7 � 0.2 1.3 � 0.3 1.3 � 0.6 2.2 � 1.0 2.0 � 0.9* 4.9 � 2.4*

Adrenaline, nmol/l a 1.1 � 0.2 1.7 � 0.3 0.8 � 0.2 1.3 � 0.2 0.8 � 0.2 1.0 � 0.1 0.7 � 0.1 0.8 � 0.2v 0.4 � 0.1 1.1 � 0.2 0.7 � 0.1 1.1 � 0.1 0.7 � 0.1 0.7 � 0.1 0.6 � 0.1 0.8 � 0.1

Values are means � SE for 11 subjects, except for plasma ATP (a, n � 10 and v, n � 8), noradrenaline (n � 6), and adrenaline (n � 6). a, Arterial; v, femoralvenous. *Different from control, P � 0.05; #different from skin hyperthermia, P � 0.05; †different from skin and mild core hyperthermia, P � 0.05. Anadditional one-way ANOVA with Tukeys honestly significant difference was performed on arterial plasma ATP values to determine any significant differencesfrom control rest conditions (§P � 0.05).

R667LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

respectively; both P � 0.05). Correspondingly, the increase inleg and systemic blood flow from rest to exercise (i.e., exercisehyperemia) declined progressively from 1.46 � 0.14 to 1.02 �0.20 l/min for LBF and 2.5 � 0.3 and 1.9 � 0.4 l/min for Qduring control compared with severe heat stress, respectively.

Circulating plasma ATP and catecholamines during wholebody heat stress. Plasma norepinephrine increased progres-sively with whole body heat stress and became significantly

elevated with Skin and Core Hyperthermia compared with bothrest and exercise control conditions (P � 0.05; Table 1).However, plasma epinephrine tended to decline throughoutwhole body heat stress at rest and during exercise (P 0.05)in association with elevations in plasma glucose and leg glu-cose uptake. At rest, venous plasma ATP remained unchangedthroughout (P 0.05; Table 1), while arterial plasma ATPincreased at rest and during exercise (P � 0.05; Table 1). This

Fig. 2. Leg hemodynamics and oxygen consumption during whole body heat stress. Leg blood flow (LBF) increased with each stage of heat stress at rest andwas also increased from control levels during exercise, while leg arteriovenous oxygen (a-vO2) difference declined; thus leg VO2 was maintained. Leg skin bloodflow (SkBF) increased at rest, while leg vascular conductance (LVC) also increased with heat stress at rest and during exercise. Data aremeans � SE for 10 subjects, except for LBF and leg SkBF, where n � 11 and n � 8, respectively. *Different from control, P � 0.05; #different from skinhyperthermia, P � 0.05; †different from skin and mild core hyperthermia. Significance refers to differences in the respective conditions, i.e., either rest orexercise.

R668 LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

increase in arterial plasma ATP was strongly correlated withincreases in LVC (r2 � 0.87; P � 0.001, Fig. 4), muscle tissuetemperature (r2 � 0.85; P � 0.001, Fig. 4, n � 5) and arterialplasma norepinephrine (r2 � 0.54; P � 0.03) at rest and during

exercise. Stepwise regression analysis revealed that the largestproportion of the variance in LVC during whole body heatstress at rest and during exercise was accounted for by eleva-tions in arterial plasma ATP (�87%).

Fig. 3. Common and deep femoral venous O2 content and estimated leg tissue blood flow with heat stress. Arterial and common femoral venous O2 content duringwhole body heat stress. Data are means � SE for 10 participants. In 4 additional participants, increases in common femoral venous O2 content were accompaniedby elevations in deep femoral venous O2 content and estimated leg tissue blood flow, which are suggestive of an increase in blood flow to the muscle tissue withinthe leg. *Different from control, P � 0.05; #different from skin hyperthermia, P � 0.05; †different from skin and mild core hyperthermia. Significance refersto differences in the respective conditions, i.e., either rest or exercise.

Table 2. Systemic hemodynamics during whole-body heat stress at rest and during exercise

Control Skin HyperthermiaSkin and Mild Core

Hyperthermia Skin and Core Hyperthermia

Rest Exercise Rest Exercise Rest Exercise Rest Exercise

Cardiac output, l/min 5.3 � 0.2 7.8 � 0.4 6.9 � 0.3* 9.2 � 0.4* 8.5 � 0.4*# 10.2 � 0.4*# 9.0 � 0.4*#† 10.9 � 0.5*#†Mean arterial pressure, mmHg 109 � 3 146 � 2 103 � 0* 129 � 2* 101 � 2*# 120 � 3*# 105 � 3† 119 � 3*#Femoral venous pressure, mmHg 11 � 1 20 � 2 11 � 2 20 � 2 14 � 2# 19 � 3 14 � 3† 18 � 3Systemic vascular conductance,

ml ·min�1 ·mmHg�1 49 � 2 54 � 3 67 � 3* 72 � 4* 84 � 4*# 86 � 5*# 86 � 4*# 91 � 4*#†Perfusion pressure, mmHg�1 97 � 3 125 � 2 90 � 2* 109 � 2* 88 � 3* 101 � 4*# 93 � 3* 100 � 3*#

Data are means � SE for 11 subjects. Cardiac output, mean arterial pressure, femoral venous pressure, and systemic vascular conductance during heat stressat rest and one-legged knee extensor exercise. *Different from control, P � 0.05; #different from skin hyperthermia, P � 0.05; †different from skin and mildcore hyperthermia. Significance was accepted at P � 0.05 and refers to differences in the respective conditions, i.e., either rest or exercise.

R669LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

Hydration, temperature, and hemodynamics during isolatedleg heat stress. After 60 min of isolated leg Skin Hyperthermia,mean leg skin temperature increased from 31.8 � 0.1°C atcontrol to 37.4 � 0.1°C (P � 0.05), while quadriceps muscletemperature also increased from 34.0 � 0.4 to 36.8 � 0.2°C(Fig. 1). However, core temperature (37.0 � 0.1°C) andhydration status, as indicated by body weight, remained un-changed. Correspondingly, LBF and muscle oxygenation in-creased from 0.47 � 0.08 to 0.96 � 0.07 l/min and 75 � 1%to 86 � 1%, respectively (P � 0.05). Furthermore, leg SkBFincreased from 17 � 0.5 to 45 � 2 arbitrary units (AU) (P �0.05), whereas Q, MAP, systemic vascular conductance, heartrate, stroke volume, and whole body VO2 all remained un-changed.

Effect of isolated leg and whole body heat stress on restingleg hemodynamics. In seven participants where leg and sys-temic hemodynamics were examined during both protocols 1and 2, whole body and isolated leg Skin Hyperthermia induceda significant but similar increase in LBF compared with control(LBF � 0.50 � 0.07 vs. 0.49 � 0.04 l/min, respectively; P �0.05, Fig. 5). However, Q only increased during whole bodyheat stress (i.e., by 2.1 � 0.3 l/min, P � 0.05). With furtherwhole body heat stress, the increase in LBF and Q doubled(i.e., 1.05 � 0.11 and 4.0 � 0.2 l/min, respectively). Thus, theincrease in LBF with isolated leg Skin Hyperthermia accountedfor up to 52 � 9% of the LBF increase observed with wholebody skin and core hyperthermia. The increases in local and

systemic perfusion were matched with an elevated leg andsystemic vascular conductance (Fig. 5).

DISCUSSION

This study reveals three key findings that provide furtherinsight into the role of local temperature on limb musclevasodilatation during heat stress and exercise hyperemia inhumans. First, leg and systemic perfusion and vascular con-ductance increased progressively with elevations in muscletissue temperature both at rest and during exercise. The in-creases in leg tissue perfusion paralleled significant reductionsin leg a-vO2 differences and O2 extraction due to increases invenous O2 content in the blood circulating through the deepand common femoral veins and the muscle microcirculation.These observations suggest that heat stress not only augmentsflow and venous O2 content in the skin vasculature but also in

Fig. 5. Systemic and local responses to whole body and isolated leg heat stress.LVC increased to a similar extent with both isolated leg and whole body skinhyperthermia in association with comparable increases in local tissue temper-ature [i.e., muscle and skin temperature (Tsk)]. Tm denotes quadriceps muscletemperature. However, systemic vascular conductance only increased withwhole body skin hyperthermia. With greater elevations in muscle and coretemperature (Tc) with severe whole body heat stress, leg and systemic vascularconductance increased further. Leg and systemic vascular conductance valuesare representative of LBF and Q, respectively. Data are means � SE for 7subjects; *different from isolated leg heat stress skin hyperthermia; #differentfrom whole body heat stress skin hyperthermia. Significance was accepted atP � 0.05.

Fig. 4. Relationship between LVC, muscle temperature, and plasma ATP. Withwhole body heat stress at rest and during exercise, LVC shares a strongpositive relationship with elevations in leg muscle tissue temperature andarterial plasma ATP (r2 � 0.87, P � 0.001). Stepwise regression analysisrevealed that the majority of the variance in LVC (87%) can be accounted forby elevations in arterial plasma ATP as opposed to the influence of muscletissue temperature. However, elevations in muscle tissue temperature shared astrong relationship with elevations in plasma ATP (r2 � 0.85 P � 0.001),which could indicate that ATP is released into the vascular system in responseto elevations in local tissue temperature, thereby initiating local vasodilatationand elevations in skeletal muscle perfusion. Data for LVC and arterial plasmaATP are means � SE for 10 subjects and 5 subjects for leg muscle tissuetemperature.

R670 LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

the skeletal muscle vasculature. Second, progressive increasein LVC, which occurred in spite of increases in plasma nor-epinephrine at rest and during exercise, was associated withelevations in muscle temperature (r2 � 0.85; P � 0.001) andarterial plasma ATP (r2 � 0.87; P � 0.001). Third, isolatedelevations in leg muscle and skin temperatures to �37°Caccounted for approximately one-half of the leg hyperemiaseen when muscle, skin, and core temperatures increasedfurther to �38–39°C during whole body heat stress. Thesefindings suggest that elevations in local tissue temperature alsoinduce vasodilatation in the leg muscle microvasculature, con-tributing to heat stress and exercise limb hyperemia.

A key question of this study was whether heat stress causedvasodilatation not only of the skin but also of the musclevasculature. To answer this question we used three differentapproaches: 1) took blood samples from the deep portion of thefemoral vein; 2) measured quadriceps tissue oxygenation withNIRS; and 3) estimated leg tissue blood flow, which excludesthe contribution of the skin circulation via the great saphenousvein. In support of the hypothesis of the study, we found thatheat stress evoked significant increases in deep femoral venousO2 content, quadriceps tissue oxygenation, and leg tissue bloodflow in parallel with significant reciprocal reductions in legtissue O2 extraction in conditions where arterial O2 content andleg VO2 remained essentially unchanged (Fig. 3). Deep femoralvenous O2 content and NIRS-determined quadriceps tissueoxygenation can be used as indexes of leg muscle oxygenationand perfusion under normal conditions. This has been shown inexperiments where increasing or reducing LBF with intrafemo-ral artery infusion of vasodilators and/or vascular signal trans-duction blockers demonstrate that changes in femoral venousO2 content or muscle tissue oxygenation largely reflect varia-tions in leg muscle blood flow in conditions of stable arterialO2 content, muscle metabolism, and environmental tempera-ture (13, 27, 36). A critical question in the present setting iswhether they also reflect increases in skeletal muscle perfusionwith exposure to heat stress, or according to the general belief,they solely mirror the increases in limb SkBF. In this respect,there are several methodological considerations needed topresent the experimental approaches.

First, the possibility exists that skin perfusion of the lowerleg contributes significantly to the increases in deep femoral O2

content, thus rendering the rise in venous O2 content with heatstress as merely a cutaneous response. In this regard, some ofthe leg SkBF can initially be drained by the small saphenousvein and thereafter flow into the deep femoral vein via anas-tomoses, thereby affecting deep femoral venous O2 contentvalues. Arguing against this possibility, the present findings ofan increase in deep venous O2 content and reciprocal reduc-tions in a-vO2 difference are consistent with the observationthat human calf blood flow is significantly elevated duringpassive leg heating (i.e., 56% increase) (21), even when themagnitude of the increase in whole LBF only representedone-third of the 1.1 l/min increase observed in the presentstudy with whole body heat stress. Second, it has been reportedthat measures of muscle tissue oxygenation using NIRS may beinfluenced by the rise in SkBF and oxygenation (7) and that themagnitude of this influence can depend on the specific NIRSsystem. In this regard, the system used here contrasts thoseused previously, as it is able to account for elevations inshallow tissue oxygenation that occur with increases in SkBF

using a two-optode system, which also emits NIRS signals thatpenetrate to a deeper tissue depth, thereby reaching further intoskeletal muscle tissue. Thus, it is quite unlikely that ourmeasures of quadriceps tissue oxygenation only reflected in-creases in SkBF in the present scenario where deep tissueoxygenation would need to be maintained or decreased (duringexercise) for leg SkBF to solely account for the observed 1.1and 0.7 l/min elevations in resting and exercising LBF withheat stress. Third, the increase in deep femoral O2 contentshown in this study was obtained in a small subgroup ofsubjects (n � 4). Although this is a limitation of the study, theconcurrent reductions in deep femoral-based leg a-vO2 differ-ence due only to increases in venous O2 content were somarked that they reached statistical significance both at restand exercise (Fig. 3). Furthermore, the significant increases indeep femoral venous O2 content at rest paralleled similarincreases in common femoral venous O2 content and quadri-ceps muscle oxygenation in all the subjects, suggesting thatthis is a general phenomenon in humans. Last, indices ofmuscle perfusion, deep femoral venous O2 content and NIRSmeasurements of quadriceps tissue oxygenation increased atrest and during exercise without a meaningful change in legmetabolism (Figs. 2 and 3). Assuming a constant leg skin VO2

we estimated that leg tissue blood flow, which excludes theblood flow through the great saphenous vein, increased by0.8–1.1 l/min in response to the significant reductions in deepfemoral-based leg a-vO2 differences evoked by heat stress atrest and during exercise (Fig. 3). Because increases in skinmetabolism would have little influence on the flow estimateand muscle is clearly the primary driver of the �0.25 l/minelevation in leg VO2 in all exercise conditions, it is reasonableto suggest that the rise in deep femoral venous O2 content andthe concomitant reductions in a-vO2 difference and O2 extrac-tion across the leg tissues with heat stress and exercise are inpart responses to the increase in leg muscle perfusion. Wepropose that the increased active muscle blood flow during heatstress exercise has an important thermoregulatory role as itremoves heat away from the exercising muscles via the circu-lation (i.e., convective heat transfer within the leg or to thebody core) (15).

The examination of the relationships between muscle, skin,and core temperature and leg perfusion during exposure toisolated leg and whole body heat stress provides a novel insightinto which temperature elevations contribute to the increases inleg tissue perfusion. A novel observation was that isolatedelevations in leg muscle and skin temperatures from 32–35°Cto �37°C (while maintaining core and other regional skintemperatures) accounted for approximately one-half of the leghyperemia seen when muscle, skin, and core temperaturesincreased further to �38–39°C during whole body heat stress.Furthermore, of all body temperatures measured, the progres-sive increases in LBF, LVC, and Q at rest and during exercisewith whole body heat stress were only significantly associatedwith elevations in quadriceps muscle temperature (r2 � 0.93–0.99 all P � 0.05). More importantly, LBF and LVC wereequally elevated when leg muscle and skin temperatures weresimilar between isolated leg and whole body heat stress pro-tocols, i.e., Skin Hyperthermia (Fig. 5), as were muscle oxy-genation and leg SkBF. These results therefore support a closerelationship between local tissue temperature and limb tissueblood flow and vascular conductance (Fig. 4), which is con-

R671LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

sistent with the rapid increases in quadriceps muscles andvenous blood temperatures during local thigh heating and/orexercise (15).

The mechanisms underpinning heat stress and local hyper-thermia induced limb muscle and skin vasodilatation are likelyto involve local vasodilator and vasoconstrictor signals andcentral neural reflexes (20, 38). Here we found that leg tissueperfusion and vascular conductance increased in the presenceof an augmented muscle sympathetic vasoconstrictor activityas evidenced by the rise in circulating norepinephrine (Table1). The observation that plasma epinephrine tended to declinesuggests that the increase in plasma norepinephrine was largelythe result of spillover of norepinephrine into the circulationfrom sympathetic nerve terminals. Studies using microneurog-raphy support this notion by showing that muscle sympatheticnerve activity is elevated with exposure to heat stress (30, 33).Our data indicates that factors related to heat stress modulatesympathetic vasoconstrictor activity such that vasodilator ac-tivity overrides vasoconstrictor activity. This prevailing vaso-dilatation resembles the functional sympatholysis occurring inskeletal muscle vasculature in conditions of increased sympa-thetic nerve drive during exercise and hypoxia. (34, 36).

A salient observation of this study was that elevations inLVC with heat stress were closely related to the rise inquadriceps muscle temperature and the increases in arterialplasma ATP at rest and during exercise (r2 � 0.85–0.87, bothP � 0.01, Fig. 4). The significant relationship implying a roleof local hyperthermia upon limb microcirculatory control isconsistent with reports from human forearm (5, 6, 19) and legstudies (24), showing that increases in local tissue temperatureare tightly related to forearm and leg vasodilatation. However,the results from studies in isolated vessel preparations areequivocal with some reports showing an altered vascular re-sponsiveness to elevations in temperature (23), while others donot (31). An alternative possibility is that local hyperthermiaacted indirectly upon vasodilatation and sympatholysis throughincreases in intravascular ATP and/or other vascular signalssensitive to temperature. In support of a potential role ofintravascular ATP, the stepwise regression analysis identifiedarterial plasma ATP as the best predictor of the increase inLVC during heat stress at rest and during exercise (r2 � 0.87,P � 0.001). Muscle temperature did not increase the predictingpower of ATP because it was significantly correlated with ATP(r2 � 0.95, P � 0.001). This could be interpreted to mean thatlocal temperature might trigger the release of ATP into thevascular lumen of the arterial tree providing the limb tissuevasculature, including that of the skeletal muscle and skin, withan augmented intravascular concentration of ATP. This is anattractive idea in light of recent observations indicating thatintravascular ATP can act as a potent vasodilator and sympa-tholytic molecule in the leg and forearm (13, 14, 22, 36, 37).Interestingly, the observed doubling in the arterial ATP con-centration could underlie a major role of ATP in heat stress-induced local and systemic hyperemia, as graded doses ofintrafemoral artery infusion of ATP, but not femoral venousinfusion, has been shown to progressively increase LBF and Qto 8 and 15 l/min, respectively (13). Therefore, the presentfindings merit further investigation into the sources and thepotential roles of intravascular ATP and local tissue tem-perature on the control of limb muscle and skin perfusionduring heat stress and exercise.

Perspectives and Significance

The concept that elevations in SkBF accounts entirely for theincreases in Q and the reductions in blood flow in the visceralorgans is a widely held dogma in human thermoregulatory andintegrative physiology (25, 39). The present and recent findings(21) challenge this classic dogma by supporting that elevationsin skeletal muscle blood flow contribute to the increases inlimb tissue and systemic perfusion with heat stress and com-bined heat stress and mild exercise. The increases in skeletalmuscle blood flow in conditions where heat stress and/orexercise markedly increase muscle sympathetic vasocontrictoractivity (30, 33) require the involvement of vasodilator andsympatholytic signals capable of overriding the augmentedneural vasoconstrictor reflexes on the muscle microvascula-ture. Our results that heat stress-mediated increases in skeletalmuscle blood flow share strong relationships with increases inmuscle tissue temperature and the arterial concentration of thevasodilator and sympatholytic molecule ATP open new ave-nues for future research investigating the signaling mecha-nisms underlying hyperthermia- and exercise-induced hyper-emia. Our knowledge and understanding of the role of muscleblood flow in thermoregulation is very limited compared withthe extensive literature in the field of muscle metabolism. Ourdata are important in this respect because they support thehypothesis that increases in skeletal muscle blood flowduring heat stress and exercise might have important ther-moregulatory implications for the control of local tissuetemperature as they can enhance the removal of heat fromthe muscles via the circulation (i.e., convective heat ex-change) (15).

ACKNOWLEDGMENTS

We thank the participants for dedication and commitment without whichthis study would not have been possible. We also acknowledge StephaneDufour, Ioannis Papanikolaou, Doireann McMorrow, and Kelly Street for helpand assistance during and in preparation for this study.

Present addresses: J. Pearson. Institute for Exercise and EnvironmentalMedicine, Texas Health Presbyterian Hospital Dallas, 7232 Greenville Ave.,Dallas, TX 75231 (e-mail: jamespearson@texashealth.org); D. A. Low, Neu-rovascular and Autonomic Medicine Unit, St. Mary’s Hospital, ClinicalNeurosciences, Faculty of Medicine, Imperial College London, London,W2 1NY, UK.

AUTHORS CONTRIBUTIONS

J. Pearson participated in the conception and design of the study, theanalysis and interpretation of data, and writing the article. D. A. Low partic-ipated in the data collection and interpretation and drafting of the article. E.Stöhr contributed to the data collection. K. Kalsi participated in the datacollection and performed the analysis of the plasma catecholamines and part ofthe plasma ATP samples. L. Ali and H. Barker participated in the datacollection and provided medical support for the study. J. González-Alonsoparticipated in the conception, design of the study, data collection, interpreta-tion of data, and drafting of the article. All authors approved the final versionof the manuscript for publication.

GRANTS

This study was funded by the Gatorade Sports Science Institute.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

REFERENCES

1. Abraham P, Leftheriotis G, Desvaux B, Saumet M, Saumet JL.Venous return in lower limb during heat stress. Am J Physiol Heart CircPhysiol 267: H1337–H1340, 1994.

R672 LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from

2. Abramson DI, Kahn A, Tuck S Jr, Turman GA, Rejal H, FleischerCJ. Relationship between a range of tissue temperature and local oxygenuptake in the human forearm. I. Changes observed under resting condi-tions. J Clin Invest 37: 1031–1038, 1958.

3. Andersen P, Adams RP, Sjogaard G, Thorboe A, Saltin B. Dynamicknee extension as model for study of isolated exercising muscle inhumans. J Appl Physiol 59: 1647–1653, 1985.

4. Barcroft H, Bonnar WM, Edholm OG. Reflex vasodilatation in humanskeletal muscle in response to heating the body. J Physiol 106: 271–278,1947.

5. Barcroft H, Edholm OG. The effect of temperature on blood flow anddeep temperature in the human forearm. J Physiol (Lond) 102: 5–20, 1943.

6. Barcroft H, Edholm OG. Temperature and blood flow in the humanforearm. J Physiol 104: 366–376, 1946.

7. Davis SL, Fadel PJ, Cui J, Thomas GD, Crandall CG. Skin blood flowinfluences near-infrared spectroscopy-derived measurements of tissue ox-ygenation during heat stress. J Appl Physiol 100: 221–224, 2006.

8. Detry JM, Brengelmann GL, Rowell LB, Wyss C. Skin and musclecomponents of forearm blood flow in directly heated resting man. J ApplPhysiol 32: 506–511, 1972.

9. Edholm OG, Fox RH, Macpherson RK. The effect of body heating onthe circulation in skin and muscle. J Physiol 134: 612–619, 1956.

10. González-Alonso J, Calbet JA. Reductions in systemic and skeletalmuscle blood flow and oxygen delivery limit maximal aerobic capacity inhumans. Circulation 107: 824–830, 2003.

11. González-Alonso J, Calbet JA, Nielsen B. Muscle blood flow is reducedwith dehydration during prolonged exercise in humans. J Physiol 513:895–905, 1998.

12. González-Alonso J, Dalsgaard MK, Osada T, Volianitis S, DawsonEA, Yoshiga CC, Secher NH. Brain and central haemodynamics andoxygenation during maximal exercise in humans. J Physiol 557: 331–342,2004.

13. González-Alonso J, Mortensen SP, Jeppesen TD, Ali L, Barker H,Damsgaard R, Secher NH, Dawson EA, Dufour SP. Haemodynamicresponses to exercise, ATP infusion and thigh compression in humans:insight into the role of muscle mechanisms on cardiovascular function. JPhysiol 586: 2405–2417, 2008.

14. González-Alonso J, Olsen DB, Saltin B. Erythrocyte and the regulationof human skeletal muscle blood flow and oxygen delivery: role ofcirculating ATP. Circ Res 91: 1046–1055, 2002.

15. González-Alonso J, Quistorff B, Krustrup P, Bangsbo J, Saltin B. Heatproduction in human skeletal muscle at the onset of intense dynamicexercise. J Physiol 524: 603–615, 2000.

16. Gorman MW, Marble DR, Ogimoto K, Feigl EO. Measurement ofadenine nucleotides in plasma. Luminescence 18: 173–181, 2003.

17. Hardy JD, Stolwijk JA. Partitional calorimetric studies of man duringexposures to thermal transients. J Appl Physiol 21: 1799–1806, 1966.

18. Homma S, Fukunaga T, Kagaya A. Influence of adipose tissue thicknesson near infrared spectroscopic signal in the measurement of humanmuscle. J Biomed Optics 1: 418–424, 1996.

19. Johnson JM, Brengelmann GL, Rowell LB. Interactions between localand reflex influences on human forearm skin blood flow. J Appl Physiol53: 744–749, 1976.

20. Johnson JM, Proppe DW. Cardiovascular adjustments to heat stress. In:Handbook of Physiology: Environmental Physiology. Bethesda, MD: Am.Physiol. Soc., 1996, sect. 4, vol. II, chapt. 11, p. 215–243.

21. Keller DM, Sander M, Stallknecht B, Crandall CG. �-Adrenergicvasoconstrictor responsiveness is preserved in the heated human leg. JPhysiol 588: 3799–3808, 2010.

22. Kirby BS, Voyles WF, Carlson RE, Dinenno FA. Graded sympatholyticeffect of exogenous ATP on postjunctional �-adrenergic vasoconstrictionin the human forearm: implications for vascular control in contractingmuscle. J Physiol 586: 4305–4316, 2008.

23. Kluess HA, Buckwalter JB, Hamann JJ, Clifford PS. Elevated temper-ature decreases sensitivity of P2X purinergic receptors in skeletal musclearteries. J Appl Physiol 99: 995–998, 2005.

24. Krustrup P, Soderlund K, Mohr M, González-Alonso J, Bangsbo J.Recruitment of fibre types and quadriceps muscle portions during re-peated, intense knee-extensor exercise in humans. Pflügers Arch 449:56–65, 2004.

25. Minson CT, Wladkowski SL, Cardell AF, Pawelczyk JA, Kenney WL.Age alters the cardiovascular response to direct passive heating. J ApplPhysiol 84: 1323–1332, 1998.

26. Mortensen SP, Damsgaard R, Dawson EA, Secher NH, González-Alonso J. Restrictions in systemic and locomotor skeletal muscle perfu-sion, oxygen supply and VO2 during high-intensity whole body exercise inhumans. J Physiol 586: 2621–2635, 2008.

27. Mortensen SP, González-Alonso J, Damsgaard R, Saltin B, HellstenY. Inhibition of nitric oxide and prostaglandins, but not endothelial-derived hyperpolarizing factors, reduces blood flow and aerobic energyturnover in the exercising human leg. J Physiol 581: 853–861, 2007.

28. Nadel ER, Cafarelli E, Roberts MF, Wenger CB. Circulatory regulationduring exercise in different ambient temperatures. J Appl Physiol 46:430–437, 1979.

29. Nielsen B, Savard G, Richter EA, Hargreaves M, Saltin B. Muscleblood flow and muscle metabolism during exercise and heat stress. J ApplPhysiol 69: 1040–1046, 1990.

30. Niimi Y, Matsukawa T, Sugiyama Y, Shamsuzzaman AS, Ito H, SobueG, Mano T. Effect of heat stress on muscle sympathetic nerve activity inhumans. J Auton Nerv Syst 63: 61–67, 1997.

31. Padilla J, Garcia-Villalon AL, Fernandez N, Monge L, Gomez B,Dieguez G. Effects of hyperthermia on contraction and dilatation of rabbitfemoral arteries. J Appl Physiol 85: 2205–2212, 1998.

32. Radegran G. Ultrasound Doppler estimates of femoral artery blood flowduring dynamic knee extensor exercise in humans. J Appl Physiol 83:1383–1388, 1997.

33. Ray CA, Gracey KH. Augmentation of exercise-induced muscle sympa-thetic nerve activity during muscle heating. J Appl Physiol 82: 1719–1725, 1997.

34. Remensnyder JP, Mitchell JH, Sarnoff SJ. Functional sympatholysisduring muscular activity. Observations on influence of carotid sinus onoxygen uptake. Circ Res 11: 370–380, 1962.

35. Roddie IC, Shepherd JT, Whelan RF. Evidence from venous oxygensaturation measurements that the increase in forearm blood flow duringbody heating is confined to the skin. J Physiol 134: 444–450, 1956.

36. Rosenmeier JB, Hansen J, González-Alonso J. Circulating ATP-in-duced vasodilatation overrides sympathetic vasoconstrictor activity inhuman skeletal muscle. J Physiol 558: 351–365, 2004.

37. Rosenmeier JB, Yegutkin GG, González-Alonso J. Activation of ATP/UTP selective receptors increase blood flow and blunt sympathetic vaso-constriction in human skeletal muscle. J Physiol 586: 4993–5002, 2008.

38. Rowell LB. Cardiovascular aspects of human thermoregulation. Circ Res52: 367–379, 1983.

39. Rowell LB. Human cardiovascular adjustments to exercise and thermalstress. Physiol Rev 54: 75–159, 1974.

40. Rowell LB, Brengelmann GL, Murray JA. Cardiovascular responses tosustained high skin temperature in resting man. J Appl Physiol 27:673–680, 1969.

41. Savard GK, Nielsen B, Laszczynska J, Larsen BE, Saltin B. Muscleblood flow is not reduced in humans during moderate exercise and heatstress. J Appl Physiol 64: 649–657, 1988.

42. Smolander J, Louhevaara V. Effect of heat stress on muscle blood flowduring dynamic handgrip exercise. Eur J Appl Physiol Occup Physiol 65:215–220, 1992.

43. Tayefeh F, Kurz A, Sessler DI, Lawson CA, Ikeda T, Marder D.Thermoregulatory vasodilation increases the venous partial pressure ofoxygen. Anesth Analg 85: 657–662, 1997.

44. van Hall G, González-Alonso J, Sacchetti M, Saltin B. Skeletal musclesubstrate metabolism during exercise: methodological considerations.Proc Nutr Soc 58: 899–912, 1999.

45. Wenger CB, Bailey RB, Roberts MF, Nadel ER. Interaction of local andreflex thermal effects in control of forearm blood flow. J Appl Physiol 58:251–257, 1985.

46. Wesseling KH, Jansen JR, Settels JJ, Schreuder JJ. Computation ofaortic flow from pressure in humans using a nonlinear, three-elementmodel. J Appl Physiol 74: 2566–2573, 1993.

47. Williams CA, Lind AR. Measurement of forearm blood flow by venousocclusion plethysmography: influence of hand blood flow during sustainedand intermittent isometric exercise. Eur J Appl Physiol Occup Physiol 42:141–149, 1979.

R673LOCAL TEMPERATURE AND LIMB MUSCLE PERFUSION

AJP-Regul Integr Comp Physiol • VOL 300 • MARCH 2011 • www.ajpregu.org

on May 16, 2011

ajpregu.physiology.orgD

ownloaded from