Mechanisms involved in heat production

This chapter struggles to remain directly relevant for Section R1(ii) of the 2023 CICM Primary Syllabus, which expects the trainees to "explain the mechanisms by which normal body temperature is maintained and regulated". Specifically, the subject was the topic of Question 6 from the second paper of 2014, which asked the candidates to "list the mechanisms involved in heat production" for some of the total 80% mark. The examiners complained that the candidates did not focus on the mechanisms of endogenous heat production in their answers. How would one summarise these mechanisms in eight minutes or less? It would have to be sufficiently detailed to score marks, but sufficiently concise to avoid a massive digression into chemical thermodynamics. Unfortunately, the author's weakness for trying to laboriously explain concepts he himself doesn't completely understand had predictably made such a digression unavoidable; but the grey box below hopefully spares the reader from the worst of these excesses.

  • Mechanisms of heat production are metabolic chemical reactions that yield heat.
    • Energy produced or consumed by a chemical reaction depends on the difference between the internal energy of the reactants and the internal energy of the products.
    • If the products have a higher standard molar entropy than the reactants, energy is liberated. 
    • Several biological reactions fall into this category, mosly involving the breaking of "high-energy" bonds.
    • For example the hydrolysis of ATP into ADP and Pi yields 30.5 kJ/mol of energy
    • This energy is used to drive endothermic chemical reactions, but it is not utilised with 100% efficiency, and the residual energy is dissipated as heat.
  • Cellular reactions which produce heat, that are quantitatively important, are mostly localised to the mitochondria, and include:
    • Oxidation of NADH
    • ATP synthesis and the activity of the electron transport chain
    • Proton leak through mitochondrial inner membrane
    • Na+/K+ ATPase activity
    • Actin/myosin cross bridge cycling
  • Sites of thermogenesis are: 
    • Organs and tissues with a high baseline metabolic rate
      • Brain (60 kJ/hr)
      • Liver (60 kJ/hr)
      • Intestine (170 kJ/hr)
      • Heart (50 kJ/hr at rest) 
    • Organs and tissues with uncoupled mitochondria or scalable metabolic activity, which can increase their metabolic rate on demand:
      • Muscle (80 kJ/hr at rest, but up to five times more with shivering, and up to twenty times more with strenuous exercise) 
      • Brown adipose tissue (from 9 to 60 kJ/100g/hr)
    • Heat which is produced by these sites is homogenised across the total body mass mostly by convection via the blood (the thermal conductivity of tissues is otherwise too low for conductive heat transfer)
  • Core thermal compartment is the well-perfused body mass (~8% of total) which contains these highly metabolically active organs where most of the heat is generated.
  • Core temperature is a highly conserved stable temperature of the core compartment, which is carefully regulated by the hypothalamus
  • Peripheral temperature of the outer layers of tissue is lower and more variable

Himms-Hagen (1976)Rolfe & Brown (1997) and Prusiner & Poe (1968) were essential for the parts of this chapter which concerned the chemistry and thermodynamics of chemical heat production. However these would be pointless side quests for the exam candidate who just wants to pass, and they are redirected to succinct reviews such as Kuht & Farmery (2014).

Heat production by chemical reactions

For this author, whose grasp of basic sciences would have never been described as anything better than "tenuous" by any of his former teachers or mentors, the handling of a basic sciences subject such as this becomes intensely anxiety-provoking, considering that the readership is often well-informed and formally educated in basic chemistry (which he himself never was). Nonetheless, what follows is an attempt to explain the basic principles of heat production in chemistry, in a way which might make sense to somebody who is completely new to the subject.

From high school science classes one can vaguely remember that exothermic reactions are those where the energy produced by forming the bonds of the products is greater than the energy required to break the bonds of the reagents. If this is the level at which one had remained for the rest of their medical career, and have moved no further, this is fine and normal. In general "chemical energy" is notoriously difficult to explain and the concept is often trampled in the rush to bring warm bodies to the clinical coalface, such that whole populations of highly educated people end up graduating to advanced degrees without a clear understanding. It would be extremely presumptuous of this author to claim a better than workmanlike grasp of it either, and so the interested reader is referred back to aforementioned high school chemistry sources and to Richard Feynman, who famously said in 1963, "It is important to realize that in physics today, we have no knowledge of what energy is".

But to make an attempt of this anyway seems important. So: all of the different manifestations of energy can be modelled by the kinetic movement of particles or the interaction of electric and magnetic fields, which mediate the interactions between those particles. "Chemical energy" is in part kinetic, insofar as it is involved in the chaotic jostling of atoms and molecules which we tend to describe as "heat" and "temperature", and in part electrical, insofar as it involves the electrostatic forces that create bonds between atoms and molecules. The interaction between these two domains is inevitable, as kinetic energy from sufficiently agitated atoms can overcome the electrostatic bonds between them and cause them to separate. Conversely, kinetic energy can be released or absorbed when new bonds are formed, as the resulting product molecule ends up being either more or less agitated, depending on what the internal arrangement of electron orbitals ends up being. And all of these purely kinetic microscopic processes can yield electrical energy by liberating charged ions, and they can absorb electrical energy by using the charged ions to form bonds (and to thereby slow themselves, or to accelerate, changing the heat energy of the system). In short, the amount of heat energy contained in a pile of atoms joined together as molecules can change when those atoms decide to rearrange themselves into different molecules.

Chemical energy is sometimes said to be "stored", which may lead to the miscommunication of the concept. It is stored only in the sense that a configuration of a chemical system (eg. a molecule) has the potential to change into another configuration (eg. two new molecules) which would have a different  total internal energy, i.e. the total kinetic energy of all its vibrating particles. The change in internal energy would of course depend on what the next configuration is. Chemical potential energy is therefore only discussed in reference to some arbitrary reference configuration of a system, and is not some absolute value. 

The difference in the thermal energy, or heat produced or absorbed by a chemical reaction, is therefore dependent on the difference between the internal energy of the reactants and the internal energy of the products. The products may have much higher internal energy, and the reaction will have absorbed heat, i.e. it would have been endothermic. Without traumatising the reader with in-depth discussions of  problematic concepts like Gibbs free energy (described as "a weed in the field of chemical thermodynamics"), it will suffice to say that all systems tend towards higher entropy and lower total energy, which means "spontaneous" reactions are usually those that produce reagents with lower internal energy and higher entropy.  The whole relationship can be described as: 



  • Δ is the free energy change,
  • ΔH  is the enthalpy of the reaction, i.e. the heat that is released or absorbed during the reaction (with pressure and volume remaining fixed and stable), in joules per kelvin per mole
  • T is the absolute temperature, and
  • ΔS  is the entropy change, also in joules per kelvin per mol, which is the sum of standard molar entropy values of products minus the sum of standard molar entropy values of reactants.

In order for the reaction to proceed spontaneously, the ΔG has to be less than 0, i.e. the free energy  must decrease, which means the TΔ must be larger than ΔH. From this it follows that if ΔS  remains stable, heat (T) must be added to the reaction in order for TΔto be larger than ΔH, producing a negative ΔG value.  Similarly, if the standard molar entropy of the products is vastly larger than the standard molar entropy of the reactants, then the TΔS will be much higher than  the Δ and the reaction is going to proceed spontaneously, producing the more stable (higher entropy) products.  In case one wishes to review the standard molar entropies of some common metabolic products, they are listed here:

Standard molar entropies of common biological products
Product     Standard Entropy (J/mol K)
H2O 69.91
CO2 213.74
Lactate 192
ATP4- 182.8
ADP3- 207.4
H2PO4 90.37

For most spontaneous biological reactions, eg. where large molecules are broken down into smaller ones or high-energy bonds are broken, the ΔS is very large. For example the standard molar entropy of the products (ADP and Pi) is much larger than the standard molar entropy of the reactants (ATP), which means that ΔG is very negative, -30.5 kJ/mol in the case of this first step of ATP hydrolysis under usual body temperature and pH conditions. 

Unless all of that that free energy is completely consumed in the process of driving some other reaction, it ends up turning into ΔH, i.e. a change in enthalpy occurs (in this case a negative change, because the ΔG is negative). A loss of enthalpy means heat is exported from the reaction, and this is the basis of heat production by chemical reactions in general. In the case of ATP hydrolysis, the Δis usually absorbed by some nearby chemical process, as ATP ends up being hydrolysed near some reagents with a positive ΔG, but it is usually not completely consumed. For example, ATP hydrolysis by the Na+/K+ ATPase is said to have a free energy of -51 kJ/mol, but the work of pumping sodium and potassium ions only requires about 38 kJ/mol, which means that about 13 kJ/mol is wasted as heat (Clarke et al, 2013).  Which finally brings us to return to something vaguely meaningful , i.e. an exploration of the chemical reactions that contribute to human heat generation.

Metabolic processes that contribute to thermogenesis

Most of the heat (probably about 90%) produced in the human body can be localised to the mitochondria, where a lot of the reactions which release heat tend to occur. Specifically, these are reactions related to the synthesis and catabolism of ATP, detailed elsewhere so we do not need to digress on them here. The interested reader is referred to the excellent work by Rolfe & Brown (1997) where the entire metabolic fate of substrates is untangled, weighed, measured, and presented to the reader with kilojoule price tags. It will suffice to borrow their calculations to list some processes here, giving some idea of where all that enthalpy is coming from. The values are in terms of kilojoules of heat energy produced per mole of O2 consumed in aerobic metabolism. 

Heat production by human metabolic processes
Metabolic process kJ/mol O2
Processes that produce ATP
Oxidation of NADH 250
Complex I 20-40
Complex III and IV 62
ATP synthase 48
Proton leak through the inner membrane 42
Processes that consume ATP
Na+/K+ ATPase 5.5
Protein synthesis 12
Actin/myosin cross bridge cycling up to 40

A short tl,dr to stimulate revision will summarise the events as follows:

  • NAD+ and FAD+ are synthesised by oxidation of NADH and FADH in a reaction that produces a large amount of heat
  •  NADH and FADH are then used reducing agents used to donate electrons to the electron transport chain. 
  • The electron transport chain uses these electrons to pump protons out of the matrix, and then lets protons leak back into the matrix through ATP synthase in order to produce ATP. 
  • The ATP synthesis is not 100% efficient because the inner mitochondrial membrane is permeable to protons, and they leak back into the matrix, bypassing ATP synthase (something like 20% leak is normal)
  • Thus, 2 of every 10 NADH oxidation reactions are "wasted", i.e. the proton leak prevents any ATP from being synthesised from the electrons donated from those reactions.
  • In this way, glucose can be "burned" to create NADH, FADH, and heat, feeding the electron transport chain without producing any ATP.
  • The leakier the membrane, i.e the more permeable to protons, the more heat is produced, and the less ATP; and specific porin proteins can be expressed to increase this permeability and enhance thermogenesis.

The temperature of mitochondria is therefore wildly higher than the rest of the cell. When Chrétien et al (2018) measured the temperature of active mitochondria in human kidney cells and fibroblasts, they used the title of their paper ("Mitochondria are physiologically maintained at close to 50 °C)" to communicate their astonishing findings. In mammals, who have extremely high metabolic rates and high body temperatures, the mitochondrial density of tissues is 4-5 times greater than in cold-blooded animals of equivalent size, and varies inversely with size (which makes sense, as smaller mammals dissipate heat more rapidly due to to their mass-to-surface-area ratio).

Sites of metabolic heat production

Though a mitochondrion seems to operate at something like 50 °C, the casual observer will note that the rest of the organism is not at 50 °C, and so clearly some heterogeneity must exist in the temperatures that can be measured in the human body. That does appear to be the case even at the cellular level, although admittedly the temperature gradient across a cell is difficult to measure, considering how temperature is conceptualised (can you really discuss the average kinetic energy in your system when the population of molecules can be counted on one hand?) Macroscopically, organs and tissues differ in their temperature, mostly on the basis of their metabolic activity, exposure to the external environment, and blood flow (which homogenises their temperature by convective heat exchange). From the above, it should follow that the sites of heat production in the body should be organs and tissues which: 

  • have a high rate of metabolism, 
  • are rich in mitochondria, 
  • have mitochondria which are highly inefficient and produce an excess of heat instead of ATP

The hottest organs are therefore predictably the brain  the liver and the heart.  This table from Ring & Houdas (1982) illustrates the range of possible measurable values from each site. The values listed are differences, in °C, from the rectal temperature of 36.85 °C.

Measurement site Minimum Maximum
Mouth -0.45 -0.3
Esophagus -0.3 -0.2
Stomach -0.2 -0.1
Liver -0.25 -0.05
Vagina -0.05 +0.05
Brain -0.25 +0.05
Nasopharynx -0.45 -0.4
Tympanic -0.4 -0.4
External auditory meatus -0.5 -0.1
Urine at the meatus -0.15 -0.1
Pulmonary artery -0.25 -0.15
Jugular vein: higher part -0.05 0
Jugular vein: lower part -0.25 -0.2
Vena cava -0.3 -0.25
Renal vein -0.25 -0.15
Coronary sinus -0.05 +0.05
Right atrium -0.25 -0.15
Right ventricle -0.2 -0.15
Left ventricle -0.25 -0.15

In this table the rectal temperature is used as the reference point, but in fact this is not a universally accepted standard. For some, the gold standard is PA catheter measurement (i.e. pulmonary artery temperature), which is even more invasive. What could be more core than pulmonary artery? Arguably, the hypothalamus itself could be considered even more central, as this is where the regulatory decisions are made on what to do with the body temperature, but practically speaking the direct measurement of hypothalamic temperature is difficult to achieve. Fortunately internal organs and tissues are all within about 0.2-0.4 °C and so the measurement of one value is usually close enough to the others, at least within the range of accuracy we need to assess fever and protect the organs from thermal damage. 

These are internal organs and tissues, one might point out. Other tissues are metabolically quieter and act as insulators, or passive heat storage masses, contributing little to the actual thermogenesis itself. This brings us to the concepts of core temperature and peripheral temperature.

Core and peripheral temperature

The exterior of the body is sufficiently different that the description of body heat inevitably encounters a two-body model which separates the human into a warm core and cooler shell. This may seem like a totally intuitive move to any modern critical care trainee, but would have probably been revolutionary in 1958 when the concept was first developed by Jürgen Aschoff and Rütger Wever. The authors borrowed from earlier works (eg. Burton & Edholm's "Man in a Cold Environment", 1955) that described how the temperature of the exterior surfaces can fluctuate depending on ambient conditions and the state of the peripheral circulation, whereas the inner body tends to remain relatively stable in the face of changing conditions, on the count of it being filled with nice warm organs. Using their original diagram and plugging in some modernised values from stereotypically obscure sources, this weird statement can be expressed graphically:

Total heat production by human organs

Remembering various values for the weight of their organs would rapidly lead the reader to realise that the heat generated by the viscera contributes 8% of the body's weight but perhaps 60-70% of the total heat production, whereas the soft tissues and resting muscle contribute relatively little, and possess considerable mass (25kg for muscle, perhaps 6-9kg for skin). Exterior layers, producing little heat themselves but having considerable specific heat and poor thermal condictivity, therefore act as insulators for the inner body, Burton and Edholm asserted. By this distinction, the "core" and "periphery" are clearly defined.

The core is then dependent on the periphery to dissipate the heat, as the core organs themselves have very little surface area; which is achieved convectively, using the circulation of blood. In this fashion the 300something kJ of energy which are produced per hour by the core organs is transported to the skin, where it is distributed over a wide surface area (say, about 1.9 m2).  The exterior is an open system in contact with the outside world, and is therefore able to transfer that heat into the environment (see the chapter on the mechanisms for heat transfer), remaining at a lower temperature than the core under most normal circumstances. This lower temperature will also obviously depend on the ambient environment and about a dozen other factors, which makes it subject to so much fluctuation that all measurements lose their meaning. Niven et al (2015) concluded that peripheral temperature measurements were usually at least 1-2 degrees different to core measurements, and are therefore untrustworthy.  Which finally brings us the most important factor for the CICM trainee, which is to mention somewhere in their answer that whereas peripheral temperature is generally lower and more variable, the core temperature is high, stable and carefully regulated.  


Kuht, James, and Andrew D. Farmery. "Body temperature and its regulation." Anaesthesia & Intensive Care Medicine 15.6 (2014): 273-278.

Romanovsky, Andrej A. Thermoregulation Part I: From basic neuroscience to clinical neurology. Elsevier, 2018.

Romanovsky, Andrej A. Thermoregulation Part II: From basic neuroscience to clinical neurology. Elsevier, 2018.

Sato, Norio. Chemical energy and exergy: an introduction to chemical thermodynamics for engineers. Elsevier, 2004.

Cooper, Melanie M., and Michael W. Klymkowsky. "The trouble with chemical energy: Why understanding bond energies requires an interdisciplinary systems approach." CBE—Life Sciences Education 12.2 (2013): 306-312.

Shapiro, Norman Zalmon, and Lloyd S. Shapley. "Mass action laws and the Gibbs free energy function." Journal of the Society for Industrial and Applied Mathematics 13.2 (1965): 353-375.

Welch, G. Rickey. "Some problems in the usage of Gibbs free energy in biochemistry.Journal of theoretical biology 114.3 (1985): 433-446.

Bent, Henry A. "The First Law. For scientists, citizens, poets and philosophers." Journal of Chemical Education 50.5 (1973): 323.

Brück, Kurt. "Heat production and temperature regulation." Perinatal physiology. Boston, MA: Springer US, 1978. 455-498.

Brück, K. "Thermal balance and the regulation of body temperature." Human physiology. Berlin, Heidelberg: Springer Berlin Heidelberg, 1989. 624-644.

Murrell, Hywel. "The Human Body II. Metabolism and Heat Regulation." Ergonomics: man in his working environment. Dordrecht: Springer Netherlands, 1965. 29-35.

Schlader, Zachary J., et al. "Orderly recruitment of thermoeffectors in resting humans." American Journal of Physiology-Regulatory, Integrative and Comparative Physiology 314.2 (2018): R171-R180.

Himms-Hagen, Jean. "Cellular thermogenesis." Annual review of physiology 38.1 (1976): 315-351.

Chrétien, Dominique, et al. "Mitochondria are physiologically maintained at close to 50 C." PLoS biology 16.1 (2018): e2003992. 

Okabe, Kohki, et al. "Intracellular temperature mapping with a fluorescent polymeric thermometer and fluorescence lifetime imaging microscopy.Nature communications 3.1 (2012): 705.

Giuliano, Karen K., et al. "Temperature measurement in critically ill orally intubated adults: a comparison of pulmonary artery core, tympanic, and oral methods." Critical care medicine 27.10 (1999): 2188-2193.

Rolfe, D. F., and Guy C. Brown. "Cellular energy utilization and molecular origin of standard metabolic rate in mammals." Physiological reviews 77.3 (1997): 731-758.

Prusiner, Stanley, and Martin Poe. "Thermodynamic considerations of mammalian thermogenesis." Nature 220.5164 (1968): 235-237.

Clarke, Ronald J., et al. "Quantitative calculation of the role of the Na+, K+-ATPase in thermogenesis." Biochimica et Biophysica Acta (BBA)-Bioenergetics 1827.10 (2013): 1205-1212.

Boerio-Goates, Juliana, et al. "Thermochemistry of adenosine." The Journal of Chemical Thermodynamics 33.8 (2001): 929-947.

Curtin, NANCY A., and ROGER C. Woledge. "Energy changes and muscular contraction." Physiological Reviews 58.3 (1978): 690-761.

Sabater, Bartolomé. "Entropy perspectives of molecular and evolutionary biology.International Journal of Molecular Sciences 23.8 (2022): 4098.

Else, P. L., and A. J. Hulbert. "Mammals: an allometric study of metabolism at tissue and mitochondrial level." American Journal of Physiology-Regulatory, Integrative and Comparative Physiology 248.4 (1985): R415-R421.

Ricquier, Daniel. "Fundamental mechanisms of thermogenesis." Comptes rendus biologies 329.8 (2006): 578-586.

Burton, Alan Chadburn, and Otto Gustaf Edholm. "Man in a cold environment. Physiological and pathological effects of exposure to low temperatures." Man in a cold environment. Physiological and pathological effects of exposure to low temperatures. (1955).