Impact of anaesthesia on thermoregulation

This chapter is not relevant to any section of the 2023 CICM Primary Syllabus, as nowhere in the syllabus document is there any mention of the influence of drugs on thermoregulation and heat exchange. But it certainly belongs in this section and there are several questions asking about it in the past exam papers:

  • Question 10 from the second paper of 2023  - impact of sedatives
  • Question 16 from the first paper of 2022 - impact of sedatives
  • Question 4 from the first paper of 2009 - antipyretic effects of paracetamol

These are sufficiently recent that we can't explain them as refugees from the ANZCA syllabus, especially considering that the ANZCA primary exam syllabus does not contain this topic either. Yet nonetheless most reasonable people would agree that specialist experts in any field that routinely sedates and anaesthetises people should have a really clear idea of what effects those measures have,

including their effects on heat exchange. Therefore:

Environmental threats to the patient's body temperature

  • ICUs and operating theatres are cooled to 18-22 ºC for comfort of staff
  • Patients are often exposed to skin and sedated to facilitate procedures
  • Air movement is often brisk (eg. 8m/min in theatres, 25 full volume exchanges per hour)
  • Air is often dry (50% humidity) which faciliates evaporative heat loss

Effects of anaesthesia on heat production

  • Reduced total body thermogenesis by about 30%; due to:
    • Reduced skeletal muscle tone and voluntary contraction
    • Reduced respiratory work (mechanical ventilation)
    • Reduced thermodgenic activity of brown adipose tissue (by volatile anaesthics)

Effect of anaesthesia on heat loss:

  • Increased conductive heat loss
    • Mostly from the core to the peripheries
    • Due to cutaneous vasodilation and blood flow redistribution
    • Vasodilated skin then acts as an excellent heat exchanger, facilitating heat loss in context of environmental influences:

    Effect of ICU and operating theatre environment on heat loss:

    • Increased radiative heat loss (70% of the total)
      • ICUs and operating theatres are cooled to 18-22 ºC for comfort of staff
      • Patients are often exposed to skin and sedated to facilitate procedures
    • Increased convective heat loss (~10-15%)
      • Air movement is often brisk (eg. 8m/min in theatres, 25 full volume exchanges per hour)
    • Increased evaporative heat loss  (~15-20%)
      • Air is often dry (50% humidity)
      • Exposed wound surfaces and body cavities facilitate evaporation by interrupting the skin barrier

    Effects of anaesthesia on thermoregulation:

    • Decreased cutaneous vasoconstriction by direct vasodilator effects of the drugs themselves as well as by decreased sympathetic activation
    • Shivering is inhibited, and completely abolished by muscle relaxants
    • Brown adipocyte thermogenesis is inhibited
    • Sweating is preserved, but occurs at a higher threshold temperature
    • Voluntary behavioural thermoregulation is made impossible

    Effect of anaesthesia on interthreshold range:

    • Normal interthreshold range: 36.7–37.1°C (i.e. 0.4 °C)
    • With anaesthesia, widens to 4.0 °C, i.e 34.9 - 38.9 °C

    Net effect of anaesthesia on heat balance:

    • Core temperature drops by 1-1.5 ºC in the first hour, mostly due to blood flow redistribution
    • Heat loss plateaus over the subsequent hours
    • Total temperature change can be up to 3-4 ºC over the first 3-4 hours

    Fortunately, there are abundant resources, many of which are free. Sessler & Todd (2000) have an excellent paper on perioperative heat balance which is sufficiently generalisable to the ICU environment. In general anything by Daniel I. Sessler seems to be gold (Sessler 2016Sessler & Mayer 2005Sessler 1993Insler & Sessler 2006), but that's a long list. 

    The ambient environment of ICU and operating theatre

    Places where humans end up being sedated and paralysed all seem designed to extract as much heat as possible from their bodies. However, this was not always so. In fact for a rather sustained time in history the tradition of the surgeons was to keep a balmy warmth in theatre (80 ºF, around 27 ºC) for reasons that seem patient-centric. Emmett Rixford (1924), writing in response to calls for cooler conditions, defended the practice by remarking that

    "a person with less than his usual clothing on lying in a room at 80° soon chills-he would chill more quickly if cold ether vapor were poured into his lungs; therefore, the patient on the operatingtable should be kept warmer than ordinarily either by artificial heat or by preventing loss of body heat."

    But over the subsequent half century the practice of "parboiling the surgeon" had become sufficiently unpopular, operations became sufficiently prolonged, and the surgeons become sufficiently important members of the room, that attention turned to their comfort.  Lord Brock (the actual Lord of Wimbledon,  but also a thoracic surgeon), writing in 1975, remarked that

    "It is dificult or impossible to achieve climatic conditions in the operating room that are acceptable to all. 

    Nor should we try to please everybody:

    "The anaesthetist is for much of the time less active than the surgeon who bears the continuing burden of the anxiety and responsibility of the operation as well as having to work in a warmer area of the room"

    The most often quoted document, a reference to which often appears at this stage, is the work by Wyon et al (1968), who invaded thirty British operating rooms in the 1960s and annoyed the working staff by asking them constantly about the "stuffiness" and measuring the beads of sweat on their brows. Their findings were that surgeons consistently preferred something close to 18-19 ºC, whereas anaesthetists and other underdressed staff were more comfortable at 21.5 ºC. The usual operating theatre temperature ranges are therefore somewhere in this range, though facility design recommendations (eg. AHFG, 2016) usually allow for a range all the way down to 16 ºC, and all the way up to 28 ºC.

    The specific preferences of intensivists, who are generally a sessile and indolent species, remain unmeasured, but we can generally expect them to remain in line with anaesthetists, as both of us spend a lot of our time sitting down and drinking coffee. Of much greater importance are the thermal requirements of ICU nursing staff, who may spend whole days performing physically demanding activities inside plastic gowns and gloves, not unlike surgeons and scrub nurses. For this reason, the temperature range in the ICU may be adjusted down to resemble something like the operating theatre. There does not seem to be any official CICM publication to dictate the exact temperature range, but the CICM sustainability document mentions that "it is desirable to keep the ICU temperature between 18 to 25 degrees".  In short, in the ICU, as in OT, the patient is likely to end up being exposed to an ambient temperature perhaps ten degrees below their thermoneutral zone

    Effect of sedation on heat production

    That being sedated and/or paralysed is going to stop you from being able to shiver seems intuitive, and most people will quickly be able to construct a series of sentences that might describe how thermoregulation is affected by anaesthesia. It is less clear what happens to the thermogenesis, specifically to the various mechanisms of nonshivering thermogenesis. As will become apparent, we mostly have unsatisfying explanations or indirect impressions for what happens there:

    • Total body oxygen consumption is decreased, which should suggest that the total body heat production should also decrease, their being inextricably linked. A part of this is the decreased oxygen consumption by decreased skeletal muscle tone, and a part of this is the loss of respiratory muscular effort, we must surely think.  La Monaca & Fodale (2012) suggest that there may also be some direct effects on mitochondrial metabolism, though in vivo studies suggest that this effect is mixed (some drugs inhibit, and some accelerate). Whatever the truth m,ay be, the magnitude of the total change is probably small (Greene, 1972, quoted figures like 88% and 84% falling from a pre-anaesthetic 100%). 
    • Brown adipocyte thermogenesis is inhibited, specifically by volatile anaesthetics. Ohlson et al (2003) found that barbiturates propofol and ketamine had absolutely no effect on this. It is a well known thing, in the sense that mainstream authors mention it (Kurz, 2001), and there is evidence that the usual noradrenergic control of this process is interrupted by volatile agents (Dicker et al, 1995), but no good mechanistic explanation in the literature. One might object that in any case this probably only plays a marginal role in the total heat balance of the adult, but in fact infants also seem to have completely absent nonshivering thermogenesis under anaesthesia (Plattner et al, 1997).

    In short, it will be clear to most people that giving somebody propofol will not immediately cause all of their mitochondria to stop working; nor does being asleep drop one's core temperature appreciably, so some heat will continue to be produced, even if there is fractionally less of it. The more important effect of anaesthesia is on the interruption of the normal mechanisms of thermoregulation.

    Effects of anaesthesia on thermoregulation

    The anaesthetised patient will, upon induction of the general anaesthetic, immediately lose all of their most important thermoregulatory mechanisms:

    • Cutaneous vasoconstriction is totally abolished by anaesthetic agents,  because of:
      • Some direct vasodilator effects, eg. of volatile agents that seem to act a bit like calcium channel blockers (Namba & Tsuchida, 1996).
      • Indirect vasodilator effects on the sympathetic nervous system, which is depressed by anaesthesia (eg. propofol may or may not have direct vasodilator effects but it sure does depress sympathetic activity)
      • Indirectly indirect effects, mediated by the inhibition of normal sympathetic reflexes that would ordinarily control the regional circulation of the skin.
    • It's not usually a problem, but if hyperthermia was the insult, then the efficiency of cutaneous heat exchange would be grossly affected, because the usual response is to increase the cardiac output and channel a large amount of it into the skin, and the cardiac output responsiveness in an anaesthetised patient is likely to be markedly depressed, again because of the aforementioned depression of autonomic reflexes.
    • Again, sweating is not usually expected from the anaesthetised patient, as that is usually the province of the surgeon, but if it were required, the patient would be still largely capable of it, albeit at a much higher threshold (Washington et al, 1993, found that it happened at 38.1 ºC instead of the pre-anaesthetic 36.6 ºC).
    • Shivering thermogenesis is inhibited modestly by sedation, markedly by anaesthesia, and totally by the use of neuromuscular junction blockers. Though the ICU patient is not usually paralysed in a sustained fashion,  under virtually every circumstance the appearance of shivering leads to the intensivists taking aggressive steps to control shivering, and so the same can be applied to this group as to the paralysed surgical patient.
    • Brown adipocyte thermogenesis is uncoupled from sympathetic control, as mentioned above.
    • So obvious that it does not need to be stated, "voluntary behavioural control" of temperature is impaired by anaesthesia, as the comatose patient is not going to complain about the air conditioning, demand blankets or threaten to leave the operating theatre to seek warm shelter elsewhere (unlike the anaesthetist, who is free to do all of those things). 

    So, we have put them in a cold environment, dropped their metabolic rate, and taken away their usual defensive mechanisms. What happens as the result is highly predictable:

    Net effect of anaesthesia on heat balance

    The overall effects of anaesthesia on heat balance can be described in this stepwise process flow:

    • Cutaneous vasoconstriction is abolished
    • The vasodilated skin becomes the recipient of increased blood flow
    • The increased blood flow rapidly equilibrates the temperature of the body core with the temperature of the surface
    • The vasodilated skin acts as an excellent heat exchange surface, resulting in increased heat loss by the four horsemen of heat exchange:
      • Radiation (a completely physics-based effect, related to the temperature difference between the temperature of the environment and the skin)
      • Conduction (into the thermally conductive surfaces around them, of which there are usually few)
      • Convection (by heating the surrounding air, which in the ICU or operating theatre moves at something like 8 m/min, with 25 air changer per hour, making it a reasonable breeze)
      • Evaporation, which for the mechanically ventilated surgical patient has less to do with respiratory water loss and more to do with evaporation of body water directly from exposed body cavity surfaces.

    Question 20 from the second paper of 2016 was asked in a way that suggested the examiners expected the candidates to remark on the relative proportional contribution of each mechanism to the total heat loss process in anaesthesia, which seems unfair, as it does not appear in any of the major texts. Stoelting is the most comprehensive, featuring a version of the graph discussed below and some points for explanation (p. 78 and 79 of the 2015 edition); Kam has literally one paragraph on this (3rd ed, p. 385), and Ganong (23rd ed) and Guyton & Hall (20th ed) have nothing. To get the answers, we must turn to primary sources such as Matsukawa et al (1995) and English et al (1991). The former experimented on human volunteers whereas the latter chose some kind of horrendous open-abdomen pig model. The findings of both papers can be summarised by the following points:

    • Conductive heat loss into surfaces is minimal as they are usually well insulated
    • Heat loss via the skin was by far the most important, accounting for about 80% of the total. Most of this occurs via radiation:
      • Radiation loss = 133W (70% of the total) 
      • Convective loss to air = 10W (~5%)
      • Skin evaporation  = 17W (~5-10%)
    • Evaporative loss from the open abdomen was the next most important mechanism (~15-20% of the total)

    Radiative heat loss under anaesthesia is not well described and must surely vary widely depending on the amount of exposed body surface area and the use of radiant heaters (to say nothing of the heat produced by overhead theatre lamps). Yes,  it is well-established physical canon and one can easily plug known values into an online calculator, but this yields a radiative heat transfer of around 670  kJ/hr, or 183 watts. Considering that the sleeping human produces probably only something like 200-300 kJ/hr, the radiative heal loss alone would rapidly cool the body down to equilibrium if this obvious overestimate was correct; and in fact it may not be such an overestimate, as real scientific sources seem to quote figures in the range of 133 W (Sessler et al, 2008). Those figures may not represent anaesthetised patients, as unfortunately most sources have strings of references that, if traced to their origin, terminate at Hardy (1938), a paper where "two normal men were studied naked under basal conditions"

    Anyway: the net effect of all these movements of heat is easily graphed, and because the graph appears in Stoelting among other resources, the exam candidates should be prepared to dutifully reproduce it and make comments on its features. These diagrams, wherever they appear, tend to all have the same generic shape, because they are all based on the same paper. Matsukawa et al, in 1995, did things to "six minimally clothed male volunteers" that resulted in these findings:

    This first graph represents the effects of the anaesthetic agents on the total body heat content, and demonstrates a marked slump that occurs in the first few hours, which was almost entirely due to an ongoing stable heat loss of around 30 kcal/hr. On the other hand, total body heat production only decreased by about 30%, and stabilised at that level over the first hour, which means it could not be held responsible for the continuing heat loss. Nor was the skin responsible for suddenly dispersing extra heat, because these minimally dressed volunteers had equilibrated to the ambient temperature of the operating theatre over the preceding three hours. The heat was being lost by redistribution, they argued, as the total amount of heat in the body decreased less than the heat in the core:


    The frequently repeated finding of this study is that heat redistribution contributed 80% to the change in temperature in the first hour after induction of anaesthesia, and 65% to the entire change in core body temperature (2.8 ºC) after the first three hours. 

    Effect of sedation on the interthreshold range

    This appeared to be an important aspect of this topic, if we go by the comments of the examiners, who mentioned "widening of hypothalamic inter-threshold range" and implored the candidates to follow with "an explanation of what this is". The interthreshold range is of course

    "the range between core temperature at the onset of shivering and that at the onset of sweating, usually between 36.5 ºC and 37.5 ºC."

    and anaesthetics do certainly widen it. The usual image that gets trotted out to demonstrate this comes from a chapter by Sessler in a 1994 edition of Millar's Anesthesia:

    wideninig of the interthreshold range from Sessler (1994)

    That interthreshold range sure is wider, and there is probably no more eloquent way to describe it graphically, which means if one forgets the numbers in the exam, one may still be able to reproduce some fragmentary version of this image and still pass. The numbers might be inconsistent and erratic across publications, but fortunately Sessler has published the majority of them anyway, and in those works the numbers are usually the same:

    • Normal interthreshold range: 36.7–37.1°C (i.e. 0.4 °C)
    • With anaesthesia, widens to 4.0 °C, i.e 34.9 - 38.9 °C

    Detail on a more granular level than this is probably not expected, but some people may be interested in the effects of individuals drugs, which is again revealed by a figure from a Sessler paper (Sessler et al, 2008). It would not be a spoiler to remark that they all do roughly the same thing, but perhaps to varying degrees:

    effects of anaesthetics on thermoregulation, from Sessler (2008)

    A much more interesting diagram would be one that plots the same regulatory thresholds against concentrations of agents that don't have CNS activity, as often the real art in the ICU is keeping somebody from doing something perfectly thermoappropriate but annoying (eg. shivering during induction of therapeutic hypothermia). But that is probably a topic for a completely different chapter, one which would have even less association with the CICM syllabus. Until such a chapter is formed into being by the powers of the author's insomnia, the reader interested in such diversions is redirected to the excellent papers by Cox & Lomax (1977)Cuddy (2004) or Ruiz (2012), though the latter may be más excesivo.


      Sessler, Daniel I., and Michael M. Todd. "Perioperative heat balance." The Journal of the American Society of Anesthesiologists 92.2 (2000): 578-578.

      Sessler, Daniel I. "Perioperative thermoregulation and heat balance." The Lancet 387.10038 (2016): 2655-2664.

      Sessler, Daniel I., David S. Warner, and Mark A. Warner. "Temperature monitoring and perioperative thermoregulation." The Journal of the American Society of Anesthesiologists 109.2 (2008): 318-338.

      Conway, Aaron. "A review of the effects of sedation on thermoregulation: insights for the cardiac catheterization laboratory." Journal of PeriAnesthesia Nursing 31.3 (2016): 226-236.

      Brock, Lord. "The importance of environmental conditions, especially temperature, in the operating room and intensive care ward." Journal of British Surgery 62.4 (1975): 253-258.

      Clark, Robert E., Louis R. Orkin, and E. A. Rovenstine. "Body temperature studies in anesthetized man: effect of environmental temperature, humidity, and anesthesia system.Journal of the American Medical Association 154.4 (1954): 311-319.

      Rixford, Emmet. "The optimum temperature of operating-rooms." California and Western Medicine 22.12 (1924): 599.

      Wyon, D. P., O. M. Lidwell, and R. E. O. Williams. "Thermal comfort during surgical operations." Epidemiology & Infection 66.2 (1968): 229-248.

      Ohlson, Kerstin BE, et al. "Thermogenesis inhibition in brown adipocytes is a specific property of volatile anesthetics." The Journal of the American Society of Anesthesiologists 98.2 (2003): 437-448.

      Kurz, A. "Effects of anaesthesia on thermoregulation." CURRENT ANAESTHESIA AND CRITICAL CARE 12.2 (2001): 72-78.

      Dicker, Andrea, et al. "Halothane selectively inhibits nonshivering thermogenesis: possible implications for thermoregulation during anesthesia of infants." The Journal of the American Society of Anesthesiologists 82.2 (1995): 491-501.

      Bindu, Barkha, Ashish Bindra, and Girija Rath. "Temperature management under general anesthesia: Compulsion or option." Journal of Anaesthesiology Clinical Pharmacology 33.3 (2017): 306-316.

      Greene, Nicholas M. "Metabolic Effects of Anesthetics." Modern Inhalation Anesthetics (1972): 271-287.

      La Monaca, Enza, and Vincenzo Fodale. "Effects of anesthetics on mitochondrial signaling and function." Current drug safety 7.2 (2012): 126-139.

      Nakamura, K., et al. "Direct vasoconstrictor and vasodilator effects of propofol in isolated dog arteries." British Journal of Anaesthesia 68.2 (1992): 193-197.

      Koch, Marc, et al. "Effects of propofol on human microcirculation." British journal of anaesthesia 101.4 (2008): 473-478.

      Namba, Hitoshi, and Hideaki Tsuchida. "Effect of volatile anesthetics with and without verapamil on intracellular activity in vascular smooth muscle." The Journal of the American Society of Anesthesiologists 84.6 (1996): 1465-1474.

      Plattner, Olga, et al. "Lack of nonshivering thermogenesis in infants anesthetized with fentanyl and propofol." The Journal of the American Society of Anesthesiologists 86.4 (1997): 772-777.

      Dunn, John C., et al. "The sweaty surgeon: raising ambient operating room temperature benefits neither patient nor surgeon." JBJS 99.6 (2017): e27.

      Washington, Denna E., et al. "Thermoregulatory responses to hyperthermia during isoflurane anesthesia in humans." Journal of applied physiology 74.1 (1993): 82-87.

      Matsukawa, Takashi, et al. "Heat flow and distribution during induction of general anesthesia." The Journal of the American Society of Anesthesiologists 82.3 (1995): 662-673.

      English, M. J. M., et al. "Heat loss in an animal experimental model." Journal of Trauma and Acute Care Surgery 31.1 (1991): 36-38.

      Hardy, James D., Eugene F. Du Bois, and G. F. Soderstrom. "Basal metabolism, radiation, convection and vaporization at temperatures of 22 to 35 C.: Six figures." The journal of nutrition 15.5 (1938): 477-497.

      DI, SESSLER. "Temperature monitoring." Anesthesia (1994): 1363-1382.

      Cox, Barry, and Peter Lomax. "Pharmacologic control of temperature regulation." Annual Review of Pharmacology and Toxicology 17.1 (1977): 341-353.

      Ruiz, Soto. "Pharmacological approaches to control of body temperature." Revista de Enfermeria (Barcelona, Spain) 35.5 (2012): 8-14.

      Cuddy, Mary Linda Stotter. "The effects of drugs on thermoregulation." AACN Advanced Critical Care 15.2 (2004): 238-253.