Clonidine decreases the vasoconstriction and shivering thresholds. It thus seems likely that the alpha2 agonist dexmedetomidine will also impair control of body temperature. Accordingly, the authors evaluated the dose-dependent effects of dexmedetomidine on the sweating, vasoconstriction, and shivering thresholds. They also measured the effects of dexmedetomidine on heart rate, blood pressures, and plasma catecholamine concentrations.


Nine male volunteers participated in this randomized, double-blind, cross-over protocol. The study drug was administered by computer-controlled infusion, targeting plasma dexmedetomidine concentrations of 0.0, 0.3, and 0.6 ng/ml. Each day, skin and core temperatures were increased to provoke sweating and then subsequently reduced to elicit vasoconstriction and shivering. Core-temperature thresholds were computed using established linear cutaneous contributions to control of sweating, vasoconstriction, and shivering. The dose-dependent effects of dexmedetomidine on thermoregulatory response thresholds were then determined using linear regression. Heart rate, arterial blood pressures, and plasma catecholamine concentrations were determined at baseline and at each threshold.


Neither dexmedetomidine concentration increased the sweating threshold from control values. In contrast, dexmedetomidine administration reduced the vasoconstriction threshold by 1.61 +/- 0.80 degrees C x ng(-1) x ml (mean +/- SD) and the shivering threshold by 2.40 +/- 0.90 degrees C x ng(-1) x ml. Hemodynamic responses and catecholamine concentrations were reduced from baseline values, but they did not differ at the two tested dexmedetomidine doses.


Dexmedetomidine markedly increased the range of temperatures not triggering thermoregulatory defenses. The drug is thus likely to promote hypothermia in a typical hospital environment; it is also likely to prove an effective treatment for shivering.

The alpha2-adrenergic agonist clonidine produces little [1] or no [2] increase in the sweating threshold (triggering core temperature). However, the drug comparably reduces the vasoconstriction and shivering thresholds. [3] These data suggest that clonidine inhibits central thermoregulatory control, as do volatile anesthetics, [4,5] propofol, [6] and opioids. [7,8] Drugs that inhibit thermoregulatory vasoconstriction [9,10] may cause core hypothermia and inhibit shivering. [9–12] Consistent with this theory, postanesthetic shivering can usually be prevented [13,14] or treated [15,16] by clonidine administration.

Dexmedetomidine is an alpha2-adrenergic agonist, [17] having approximately 10 times the receptor specificity of clonidine. [18] It is thus likely that dexmedetomidine also produces central thermoregulatory inhibition, promotes hypothermia, and will be an effective treatment for shivering. The thermoregulatory effects of dexmedetomidine, however, have yet to be reported in humans. Accordingly, we evaluated the dose-dependent effects of dexmedetomidine on the sweating, vasoconstriction, and shivering thresholds.

The sympatholytic effects of dexmedetomidine attenuate hyperadrenergic responses to perioperative stress. [19] Extremes of temperature, particularly shivering, are also associated with increases in heart rate, blood pressure, circulating catecholamine concentrations, and oxygen consumption. [20,21] We therefore postulated that dexmedetomidine might attenuate the hyperadrenergic stress response to shivering, once triggered. Accordingly, we evaluated heart rate, blood pressures, and plasma catecholamine concentrations at each major thermoregulatory response threshold.

With approval from the Institutional Review Board of the University of California, San Francisco, and written informed consent, we studied nine male volunteers having the following morphometric characteristics: age, 28 +/- 5 yr (mean +/- SD); height, 163 +/- 24 cm; and weight, 74 +/- 11 kg. The study excluded volunteers who had a history of thyroid disease, dysautonomia, Raynaud's syndrome, and hepatic or renal disease. We also excluded those who smoked, who had a history of alcohol or drug abuse, who were taking medication, or whose body weight exceeded 115% of normal.

The study was a randomized, double-blind, cross-over comparison of two doses of dexmedetomidine and placebo (normal saline). At least 7 days were allowed between treatments. On each study day, volunteers were warmed until sweating was observed and then gradually cooled until vasoconstriction and shivering occurred. Core-temperature response thresholds were then determined by arithmetically compensating for alterations in skin temperature using previously determined cutaneous contributions to thermoregulatory control.

Treatment Protocol 

Studies started at approximately 8:00 AM, and the volunteers fasted 8 h before arriving at the laboratory. They were minimally clothed and rested supine in a 22–23 [degree sign] Celsius room during the protocol. Studies were scheduled so that thermoregulatory responses were triggered at similar times each day to minimize circadian fluctuations.

Catheters were inserted in a hand vein for fluid and study drug administration and in a radial artery for blood pressure measurement. Five ml/kg of lactated Ringer's solution was administered before study drug administration; Ringer's solution was subsequently infused at a rate of 1.5 ml [center dot] kg sup -1 [center dot] h sup -1. Volunteers were randomly assigned to each receive a continuous intravenous infusion of one of two doses of dexmedetomidine or placebo starting 1 h before thermal manipulations. The study drug was administered using a computer-controlled infusion pump, targeting plasma dexmedetomidine concentrations of 0.3 and 0.6 ng/ml. The infusion pump (Harvard Apparatus 22, Harvard Apparatus, South Natick, MA) was controlled using STANPUMP software (generously provided by Steven Shafer, M.D., Department of Anesthesia, Stanford University); the software adjusted the infusion rate every 10 s, as necessary, based on pharmacokinetic data for dexmedetomidine.

Throughout the protocol, arms were protected from active warming and cooling to avoid locally mediated vasomotion. [22] However, all other skin below the neck was similarly manipulated. Skin and core temperatures were first gradually increased with a forced-air warmer (Augustine Medical, Inc., Eden Prairie, MN) and circulating-water mattress (Cincinnati Sub-Zero, Cincinnati, OH) until significant sweating was achieved. Skin and core temperatures were then gradually decreased, using the circulating-water mattress and a prototype forced-air cooler (Augustine Medical, Inc.). [23] We took care throughout the protocol to minimally stimulate the volunteers. Temperature changes were restricted to <or= to 3 [degree sign] Celsius/h because this rate is unlikely to trigger dynamic thermoregulatory responses, [24] and the transition from cutaneous warming to cooling was made gradually. The study ended each day when shivering was detected.


Core temperature was recorded from the tympanic membrane. Thermocouple probes (Mallinckrodt Anesthesiology Products, Inc., St. Louis, MO) were inserted until volunteers felt the tip contact the tympanic membrane; the aural canal was then occluded with cotton, and a gauze pad was taped over the ear. Mean skin-surface temperature was calculated from measurements at 15 area-weighted sites. [25] Temperatures were recorded at 5-min intervals from thermocouples connected to Iso-Thermex[registered sign] thermometers having an accuracy of 0.1 [degree sign] Celsius (Columbus Instruments, Corp., Columbus, OH).

Sweating was continuously quantified on the left upper chest using a ventilated capsule. [26] Sweating rate exceeding 40 g [center dot] m sup -2 [center dot] h sup -1 for at least 5 min defined sweating threshold. [6] Absolute right middle fingertip blood flow was quantified using venous-occlusion volume plethysmography at 5-min intervals. [27] A decrease in fingertip blood flow for at least 5 min, as determined by a blinded observer, defined the vasoconstriction threshold. Shivering was evaluated using oxygen consumption, as measured by a metabolic monitor (Deltatrac[trademark symbol], SensorMedics Corp., Yorba Linda, CA). The system was used in the canopy mode, and measurements were recorded at 1-min intervals. An increase in oxygen consumption for at least 5 min, as determined by a blinded observer, defined the shivering threshold.

Arterial blood pressure (systolic, diastolic, and mean) and heart rate were measured continuously (Propaq 106, Protocol Systems, Beaverton, OR) from 5 min before until 3 h after the end of study drug infusion. Arterial blood pressure was measured via a radial artery cannula connected to a Transpac II transducer (Abbott Laboratories, North Chicago, IL). Hemoglobin oxygen saturation was measured by a pulse oximeter incorporated into the Propaq 106. Hemodynamic data and oxygen saturation were determined at 10-s intervals, and median values determined over 3-min epochs.

Dexmedetomidine, norepinephrine, and epinephrine plasma concentrations were determined from arterial blood samples collected just before the dexmedetomidine infusion was started and at the sweating, vasoconstriction, and shivering thresholds. Blood samples were immediately placed on ice and subsequently separated in a refrigerated centrifuge; plasma samples were stored at -70 [degree sign] Celsius until analysis.

Dexmedetomidine concentrations were assayed at Abbott Laboratories by gas chromatography and mass spectrometry. The method has a lower limit of detection of 20 pg/ml and coefficient of variation of 5.7% in the relevant concentration range (personal communication, Abbott Laboratories). Concentrations of epinephrine and norepinephrine in plasma were determined using high-pressure liquid chromatography, with coulometric electrochemical detection. The method has a lower limit of detection of 20 pg/ml and coefficients of variation near 10% in the relevant concentration ranges.

Data Analysis 

The cutaneous contribution to sweating [28] and to vasoconstriction and shivering [29] is linear. We thus used measured skin and core temperatures in [degree sign] Celsius at each threshold to calculate the core-temperature threshold that would have been observed had skin been maintained at one designated temperature:Equation 1where the fractional contribution of mean skin temperature to the threshold was termed beta. TCore(calculated) thus equals the measured core temperature, TCore, plus a small correction factor consisting of beta/(1 - beta) multiplied by the difference between actual (Tskin) and designated [Tskin(designated] skin temperatures. We have previously described the derivation, validation, and limitations of this equation. [6] We used a beta of 0.1 for sweating [28] and a beta of 0.2 for vasoconstriction and shivering. [29] The designated skin temperature was set at 34 [degree sign] Celsius, a typical intraoperative value.

Response thresholds, the sweating-to-vasoconstriction interthreshold range (temperatures not triggering thermoregulatory defenses), the vasoconstriction-to-shivering range, hemodynamic responses, and catecholamine concentrations at each drug dose were compared using repeated-measures analysis of variance (ANOVA) and Scheffe's F tests. The effect of treatment order was similarly evaluated. Dose-dependence of the thermoregulatory responses was determined using linear regression in individual volunteers. The resulting slopes were then compared using repeated-measures ANOVA and Scheffe's F tests. All results are presented as mean +/- SD; P < 0.01 was considered statistically significant.

Plasma dexmedetomidine concentrations exceeded target values (Table 1). Volunteers were typically mildly sedated during low-dose dexmedetomidine administration ([approximately] 0.4 ng/ml). In contrast, they were deeply sedated, but arousable, at the higher dose ([approximately] 0.8 ng/ml). Plasma dexmedetomidine concentrations were similar at each thermoregulatory threshold, indicating that disposition of the drug was not temperature-dependent. Treatment order did not confound the results. The skin and core temperatures at the sweating, vasoconstriction, and shivering are shown in Table 1, along with the calculated core-temperature thresholds.

Inspection of the individual results showed that the concentration-response curves were linear. Neither dexmedetomidine concentration significantly increased the sweating threshold from control values. In contrast, dexmedetomidine administration reduced the vasoconstriction threshold by 1.6 +/- 0.8 [degree sign] Celsius [center dot] ng sup -1 [center dot] ml, r2= 0.88 +/- 0.19. Dexmedetomidine administration similarly decreased the shivering threshold by 2.4 +/- 0.9 [degree sign] Celsius [center dot] ng sup -1 [center dot] ml, r2= 0.93 +/- 0.10. Reductions in the vasoconstriction and shivering thresholds each differed significantly from sweating, but not from each other (Figure 1).

The sweating-to-vasoconstriction interthreshold range increased from 0.1 +/- 0.2 [degree sign] Celsius during saline administration to 0.5 +/- 0.4 [degree sign] Celsius during low-dose dexmedetomidine, and then to 1.3 +/- 0.7 [degree sign] Celsius with high-dose dexmedetomidine. The vasoconstriction-to-shivering range, however, did not increase significantly with either dexmedetomidine dose (Figure 2).

Sweating increased the heart rate [approximately] 25% from baseline values during saline administration. During low- and high-dose dexmedetomidine administration, sweating had little effect on heart rate. Systolic blood pressure was reduced by sweating in all subjects, but the decrease was significantly greater during dexmedetomidine administration. At the vasoconstriction threshold, heart rate and blood pressure decreased with both doses of dexmedetomidine but did not differ significantly from baseline with saline. At the shivering threshold, heart rate increased [approximately] 18% and blood pressure increased [approximately] 12% during saline administration, but did not change significantly with either dose of dexmedetomidine.

Plasma norepinephrine concentrations were less than baseline values with both doses of dexmedetomidine at the sweating and vasoconstriction thresholds, but did not differ significantly from baseline at the shivering threshold. During saline administration, plasma norepinephrine and epinephrine concentrations increased at shivering threshold (Table 2). Plasma epinephrine concentrations were below baseline with both doses of dexmedetomidine at all thresholds (Table 2).

Dexmedetomidine was administered to target one of the highest plasma concentrations that has been used for investigations (0.6 ng/ml) and half that concentration (which is known to attenuate postoperative tachycardia). [30] Actual plasma concentrations exceeded these target values (Table 1). Dexmedetomidine plasma concentrations near 0.8 ng/ml produced considerably greater inhibition of cold responses than 75 micro gram clonidine ([approximately] 1.5 vs. 0.5 [degree sign] Celsius)[3] and even more than clonidine 8 micro gram/kg. [2] This increased effect surely results largely from the relatively high doses of dexmedetomidine that we administered. However, it may also result in part from greater alpha2specificity of dexmedetomidine. [18]

Dexmedetomidine administration decreased the vasoconstriction and shivering thresholds by similar amounts. General anesthetics and alfentanil also comparably reduce the two major cold-defense thresholds; this pattern seems more likely to result from a central than peripheral mechanism. In contrast, meperidine inhibits shivering twice as much as vasoconstriction. This special antishivering action of meperidine appears to result from stimulation of kappa-opioid receptors [31] that are primarily located in the spinal cord. Just as clonidine is an effective treatment for shivering, [13–16] the observed threshold reductions during dexmedetomidine administration suggest that it also is likely to be effective. However, dexmedetomidine does not appear to posses special antishivering properties like meperidine, which inhibits shivering more than vasoconstriction. [8]

All anesthetics and sedatives that substantially reduce cold-response thresholds also significantly increase the sweating threshold. [4–8] Clonidine, however, has little [1] if any [2] effect on sweating. Consistent with these observations, the sweating threshold remained unchanged during dexmedetomidine administration. The interthreshold range, nonetheless, increased from its normal value near 0.2 [degree sign] Celsius [24] to 0.5 [degree sign] Celsius and then to 1.3 [degree sign] Celsius during administration of the low and high dexmedetomidine doses. Dexmedetomidine thus markedly increased the range of temperatures not triggering thermoregulatory defenses.

Dexmedetomidine decreases in heart rate, blood pressures, and circulating catecholamine concentrations in healthy volunteers [32,33] and surgical patients. [19] We similarly observed that heart rate was reduced 10 beats/min and that systolic blood pressure was 20 mmHg less during dexmedetomidine than saline administration. The hemodynamic effects of dexmedetomidine were comparable at each tested dose. Plasma concentrations near 0.4 ng/ml thus appear to produce maximal hemodynamic effects in young, healthy volunteers. In distinct contrast, the vasoconstriction and shivering thresholds increased linearly with dose over the entire tested range. The hemodynamic and thermoregulatory consequences of dexmedetomidine thus appear to be mediated differently.

Based on hemodynamic responses and plasma catecholamine concentrations, shivering appeared to be the only portion of the study that produced substantial autonomic activation. At the sweating and vasoconstriction thresholds, dexmedetomidine decreased plasma norepinephrine concentrations from 111 +/- 27 pg/ml to 26 +/- 10 pg/ml and from 123 +/- 49 pg/ml to 24 +/- 7 pg/ml, respectively. In response to shivering, however, norepinephrine levels, heart rate, and systolic blood pressures returned to baseline values. These findings suggest that sufficient stress can still activate autonomic defenses, even at the relatively high dexmedetomidine concentrations used in this study.

In addition to the threshold, a thermoregulatory response is characterized by gain (incremental increase in response with further deviation in core temperature) and maximum response intensity. The gain of sweating, for example, remains normal during isoflurane [34] and enflurane [26] anesthesia, and during clonidine administration. [1] However, the gain of vasoconstriction is reduced by desflurane, [35] and the gain of shivering may also be slightly reduced by nitrous oxide. [36] The effects of dexmedetomidine on vasoconstriction, especially, may be complicated because of the drug's combined central and peripheral actions (although central effects predominate at the plasma concentrations we used [32]). Most likely, the observed reduction in the vasoconstriction threshold results from a central action of dexmedetomidine. However, we did not evaluate gain or maximum intensity in this protocol, and therefore cannot rule out the possibility that peripheral stimulation of alpha2receptors on arteriovenous shunts [37] might increase the gain of this response, once triggered.

In summary, dexmedetomidine markedly increased the range of temperatures not triggering thermoregulatory defenses because of it decreases the vasoconstriction and shivering thresholds and does not change the sweating threshold. The drug is thus likely to promote hypothermia in a typical hospital environment; it is also likely to be an effective treatment for shivering. The concentration-response relationship for thermoregulatory thresholds but not for hemodynamic or catecholamine effects were linear. These data imply that the thermoregulatory and hemodynamic effects of dexmedetomidine are mediated differently.

Delaunay L, Herail T, Sessler DI, Lienhart A, Bonnet F: Clonidine increases the sweating threshold, but does not reduce the gain of sweating. Anesth Analg 1996; 83:844-8.
Nicolaou G, Chen AA, Johnston CE, Kenny GP, Bristow GK, Giesbrecht GG: Clonidine decreases vasoconstriction and shivering thresholds, without sweating thresholds. Can J Anaesth, 1997; 44(6):636-42.
Delaunay L, Bonnet F, Liu N, Beydon L, Catoire P, Sessler DI: Clonidine comparably decreases the thermoregulatory thresholds for vasoconstriction and shivering in humans. Anesthesiology 1993; 79:470-4.
Xiong J, Kurz A, Sessler DI, Plattner O, Christensen RK, Dechert M, Ikeda T: Isoflurane produces marked and non-linear decreases in the vasoconstriction and shivering thresholds. Anesthesiology 1996; 85:240-5.
Annadata RS, Sessler DI, Tayefeh F, Kurz A, Dechert M: Desflurane slightly increases the sweating threshold, but produces marked and non-linear decreases in the vasoconstriction and shivering thresholds. Anesthesiology 1995; 83:1205-11.
Matsukawa T, Kurz A, Sessler DI, Bjorksten AR, Merrifield B, Cheng C: Propofol linearly reduces the vasoconstriction and shivering thresholds. Anesthesiology 1995; 82:1169-80.
Kurz A, Go JC, Sessler DI, Kaer K, Larson M, Bjorksten AR: Alfentanil slightly increases the sweating threshold and markedly reduces the vasoconstriction and shivering thresholds. Anesthesiology 1995; 83:293-9.
Kurz A, Ikeda T, Sessler DI, Larson M, Bjorksten AR, Dechert M, Christensen R: Meperidine decreases the shivering threshold twice as much as the vasoconstriction threshold. Anesthesiology 1997; 86(5):1046-54.
Frank SM, Shir Y, Raja SN, Fleisher LA, Beattie C: Core hypothermia and skin-surface temperature gradients. Anesthesiology 1994; 80:502-8.
Just B, Trevien V, Delva E, Lienhart A: Prevention of intraoperative hypothermia by preoperative skin-surface warming. Anesthesiology 1993; 79:214-8.
Just B, Delva E, Camus Y, Lienhart A: Oxygen uptake during recovery following naloxone. Anesthesiology 1992; 76:60-4.
Sessler DI, Rubinstein EH, Moayeri A: Physiological responses to mild perianesthetic hypothermia in humans. Anesthesiology 1991; 75:594-610.
Quintin L, Roudot F, Roux C, Macquin I, Basmaciogullari A, Guyene T, Baubourdolle M, Vaile JP, Bonnet F, Ghignone M: Effect of clonidine on the circulation and vasoactive hormones after aortic surgery. Br J Anaesth 1991; 66:108-15.
Delaunay L, Bonnet F, Duvaldestin P: Clonidine decreases postoperative oxygen consumption in patients recovering from general anaesthesia. Br J Anaesth 1991; 67:397-401.
Joris J, Banache M, Bonnet F, Sessler DI, Lamy M: Clonidine and ketanserin both are effective treatments for postanesthetic shivering. Anesthesiology 1993; 79:532-9.
Quintin L, Viale JP, Annat G, Hoen JP, Butlin E, Cottet-Emard JM, Levron JC, Bussery D, Motin J: Oxygen uptake after major abdominal surgery: Effect of clonidine. Anesthesiology 1991; 74:236-41.
Doze VA, Chen BX, Maze M: Dexmedetomidine produces a hypnotic-anesthetic action in rats via activation of central alpha-2 adrenoceptors. Anesthesiology 1989; 71:75-9.
Virtanen R, Savola JM, Saano V, Nyman L: Characterization of the selectivity, specificity and potency of medetomidine as an alpha-2 adrenoceptor agonist. Eur J Pharmacol 1988; 150:9-14.
Aho M, Scheinin M, Lehtinen AM, Erkola O, Vuorinen J, Korttila K: Intramuscularly administered dexmedetomidine attenuates hemodynamic and stress hormone responses to gynecologic laparoscopy. Anesth Analg 1992; 75:932-9.
Kurz A, Sessler DI, Narzt E, Bekar A, Lenhardt R, Huemer G: Postoperative hemodynamic and thermoregulatory consequences of intraoperative core hypothermia. J Clin Anesth 1995; 7:359-66.
Frank SM, Higgins MS, Breslow MJ, Fleisher LA, Gorman RB, Sitzmann JV, Raff H, Beattie C: The catecholamine, cortisol, and hemodynamic responses to mild perioperative hypothermia. Anesthesiology 1995; 82:83-93.
Wurster RD, McCook RD: Influence of rate of change in skin temperature on sweating. J Appl Physiol 1969; 27:237-40.
Kurz A, Sessler DI, Birnbauer F, Illievich U, Spiss C: Thermoregulatory vasoconstriction impairs active core cooling. Anesthesiology 1995; 82:870-6.
Lopez M, Sessler DI, Walter K, Emerick T, Ozaki M: Rate and gender dependence of the sweating, vasoconstriction, and shivering thresholds in humans. Anesthesiology 1994; 80:780-8.
Sessler DI, McGuire J, Moayeri A, Hynson J: Isoflurane-induced vasodilation minimally increases cutaneous heat loss. Anesthesiology 1991; 74:226-32.
Lopez M, Ozaki M, Sessler DI, Valdes M: Physiological responses to hyperthermia during epidural anesthesia and combined epidural/enflurane anesthesia in women. Anesthesiology 1993; 78:1046-54.
Rubinstein EH, Sessler DI: Skin-surface temperature gradients correlate with fingertip blood flow in humans. Anesthesiology 1990; 73:541-5.
Nadel ER, Metchell JW, Stolwijk JAJ: Control of local and total sweating during exercise transients. Int J Biometeorol 1971; 15:201-6.
Cheng C, Matsukawa T, Sessler DI, Kurz A, Merrifield B, Lin H, Olofsson P: Increasing mean skin temperature linearly reduces the core-temperature thresholds for vasoconstriction and shivering thresholds in humans. Anesthesiology 1995; 82:1160-8.
Talke P, Li J, Jain U, Leung J, Drasner K, Hollenberg M, Mangano DT: Effects of perioperative dexmededetomidine infusion in patients undergoing vascular surgery. Anesthesiology 1995; 82:620-33.
Kurz M, Belani K, Sessler DI, Kurz A, Larson M, Blanchard D, Schroeder M: Naloxone, meperidine, and shivering. Anesthesiology 1993; 79:1193-201.
Bloor BC, Ward DS, Belleville JP, Maze M: Effects of intravenous dexmedetomidine in humans. Anesthesiology 1992; 77:1134-42.
Kallio A, Scheinin M, Koulu M, Ponkilainen R, Ruskoaho H, Viinamaki O, Scheinin H: Effects of dexmedetomidine, a selective alpha-2 adrenoceptor agonist, on hemodynamic control mechanisims. Clin Pharmacol Ther 1989; 46:33-42.
Washington D, Sessler DI, Moayeri A, Merrifield B, Prager M, McGuire J, Belani K, Hudson S, Schroeder M: Thermoregulatory responses to hyperthermia during isoflurane anesthesia in humans. J Appl Physiol 1993; 74:82-7.
Kurz A, Xiong J, Sessler DI, Dechert M, Noyes K, Belani K: Desflurane reduces the gain of thermoregulatory arterio-venous shunt vasoconstriction in humans. Anesthesiology 1995; 83:1212-9.
Passias TC, Mekjavic IB, Eiken O: The effect of 30% nitrous oxide on thermoregulatory responses in humans during hypothermia. Anesthesiology 1992; 76:550-9.
Hales JRS: Skin arteriovenous anastomoses, their control and role in thermoregulation, Cardiovascular Shunts: Phylogenetic, Ontogenetic and Clinical Aspects, Edited by Johansen K, Burggren W. Copenhagen, Munksgaard, 1985, pp 433-51.