Physiology · Unit 8: Defense, skin and temperature
Skin, Sweat and Human Thermoregulation
How does a sheet of tissue a few millimetres thick keep water in, let heat out at a controlled rate and start a hormone pathway with sunlight? This lesson builds human skin from its layers, locates its water barrier with a classic tracer experiment in neonatal mice, writes the whole-body heat-balance equation and follows the human effectors in heat and cold. Afterwards you can calculate the sweat rate that balances a heat load and the rate of heat storage when humid air caps evaporation, and predict which variables rise and fall.
Watch first Sweat Cools Only When It Evaporates · 2:29
Read the transcript of “Sweat Cools Only When It Evaporates”
Film transcript: Sweat Cools Only When It Evaporates
A gram of sweat that evaporates from your skin carries off about 2.4 kJ of heat. A gram that drips off carries away almost none. So the cooling comes from evaporation, not from sweat itself. Sweat cools you only as fast as air takes up water vapour. On a humid day, that can lag behind the heat your muscles make. For a person, heat balance is a sum. Heat storage equals metabolic rate, minus external work, plus or minus dry exchange, minus evaporative loss. Dry exchange means radiant, convective and conductive heat, and it can run either way. Every term is in watts. When storage is zero, body temperature is steady. Take illustrative numbers for a healthy 70 kg adult cycling steadily in hot, sunny air. Metabolic rate is 1,000 W, and 20 percent of it becomes external work.
That leaves 800 W of heat. Dry heat gain adds 50 W, and respiratory heat loss is ignored. To keep storage at zero, evaporation must remove 850 W. Evaporating sweat takes up about 2.43 kJ per g at skin temperature. So 850 W needs about 1.26 L of sweat per hour, if every gram evaporates. Suppose only 70 percent evaporates and the rest drips. The sweat rate then rises to about 1.80 L per hour. Without drinking, body mass falls about 2.6 percent in that hour. Now let humid air cap evaporation at 700 W, when the body needs 850 The gap of 150 W is stored as heat, 540 kJ in an hour. Take 2.98 kJ per kg per degree C. Then 70 kg stores about 209 kJ per degree.
If body temperature is uniform, it rises about 2.6 degrees C per hour. Once evaporation reaches the most that air allows, sweating harder adds dripping, not cooling. Balance returns three ways. Make less heat by slowing down. Take in less dry heat with shade or cooler air. Or have drier air. These numbers are made up for practice, and the rate predicts no real person. Keep this rule. Sweat cools only when it evaporates, and the air sets how much can.
A gram of sweat that evaporates from your skin carries off about 2.4 kJ of heat. A gram that drips off carries away almost none. Sweat cools you only as fast as air takes up water vapour, which on a humid day can lag behind the heat muscles make. This lesson builds the skin, finds its water barrier, follows heat through it, and ends with vitamin D.
The four routes of heat exchange, thermal conductance and the comparative thermoneutral zone are in metabolic rate and body temperature. Here they get human numbers.
Skin is a layered barrier that renews from below
Skin has three compartments. The epidermis, a stratified epithelium, is thinnest on the eyelids, about 0.05 mm, and thickest on the palms and soles, around half a millimetre or more. The dermis is connective tissue; its deeper, reticular part holds hair roots, glands, vessels and nerves. The hypodermis anchors the dermis to muscle and bone and is mostly white fat, an insulating layer.
The epidermis is a conveyor belt. Keratinocytes divide in the stratum basale and move outward through the spinosum, the granulosum, a clear lucidum found only in the thick skin of palms and soles, and the corneum. On the way they fill with keratin and die, ending as corneocytes: flat cells with no nucleus, packed with keratins 1 and 10. Repeated friction thickens the stratum corneum into a callus; nails and hair are hard, keratinized products of specialized epidermis.
Melanocytes sit in the basal layer and make melanin, which absorbs ultraviolet light; each supplies about 30 to 40 keratinocytes. Skin colour differs with the amount and type of melanin and the size, distribution and fate of melanosomes, not with melanocyte number, which is similar between people. Langerhans cells, antigen-presenting cells most prominent in the spinosum, carry antigen to lymph nodes (lymphoid organs and immune traffic); in mice they are seeded before birth and renew locally.
The epidermis has no blood or lymph vessels and is fed by diffusion, so the vessels that open in heat are dermal. Small arrector pili muscles raise hairs in the cold; in sparsely haired humans the goosebumps trap little air, and the response is usually called vestigial.
The water barrier is lipid between dead cells
Gut and kidney epithelia seal the spaces between cells with tight junctions (epithelia), so the natural guess was that skin does too.
In 1975 Peter Elias and Daniel Friend tested this in the epidermis of neonatal mice. They introduced water-soluble tracers that are dense under the electron microscope, horseradish peroxidase and lanthanum among them, followed them between cells, and examined freeze-fracture replicas of the upper layers. If tight junctions between living cells were the seal, the tracers should have stopped there. Instead they moved between cells up to the upper stratum granulosum and stopped where lamellar bodies, lipid-filled granules, had emptied their contents into the intercellular space. The authors placed the main barrier in that extruded lipid, not in tight junctions.
Later work resolved the lipid as patterned lamellae between corneocytes, often pictured as bricks in mortar. Then tight junctions returned: Furuse and colleagues (2002) found that mice lacking claudin-1 died within a day of birth with excess water loss through the skin, so granular-layer tight junctions also contribute. Both now count.
Heat balance: what the skin circulation controls
For a person, the whole-body heat balance is:
S = (M - W) ± R ± C ± K - E
S is heat storage, M metabolic rate, W external work, R, C and K radiant, convective and conductive (dry) exchange, which runs either way, and E evaporative loss, all in watts. When S is zero, body temperature is steady.
At steady state the heat that reaches the skin from the core, carried mostly by blood, equals the heat the skin loses, so the skin circulation acts as a variable insulator. In a model of an average man, total body insulation falls about four-fold, from about 0.124 to 0.031 m2 degrees C per W, as skin blood flow goes from minimal to maximal.
The human thermoneutral zone is the range of ambient temperature over which body temperature is regulated by dry heat loss alone, through skin blood flow, with no regulatory change in heat production or evaporation. Below its lower limit heat production must rise; above its upper limit regulatory sweating begins. For resting nude people in air, Hardy and DuBois reported a lower critical temperature of about 28.5 degrees C, with skin at 33.5 degrees C; Kingma and colleagues (2014), reporting that work, argued that skin temperature belongs in the definition.
Transmitters and directions:
- Vasoconstriction in the cold: sympathetic nerves release noradrenaline onto skin arterioles.
- Active vasodilation in heat, in hairy skin: sympathetic cholinergic nerves, with cotransmitters and nitric oxide; no single mediator has been identified. The system has no resting tone and switches on only as body temperature rises, yet supplies 80 to 90 percent of the vasodilation in hairy skin in whole-body heating. Withdrawal of constrictor tone supplies the rest. The glabrous skin of the palms and soles has no active vasodilator system; there, warming dilates vessels mainly by withdrawing constrictor tone.
- In hyperthermia, skin blood flow can reach 6 to 8 L per min, a load on cardiac output and venous return.
TRP-channel sensing is in thermosensation, the preoptic integrator in the hypothalamus and autonomic control. The regulated temperature is a range: in 148 healthy adults aged 18 to 40, mean oral temperature was 36.8 degrees C (Mackowiak, Wasserman and Levine, 1992), not Wunderlich’s 37 degrees C, which came from axillary readings.
Sweat cools only when it evaporates
Humans have about 2 to 4 million eccrine sweat glands, densest on the palms and soles at about 250 to 550 per cm2. Their total number is fixed by about age 2 to 3, so density falls as the body grows. Sweating responds mainly to core temperature, with a skin contribution, through the preoptic area. Sympathetic postganglionic fibres release acetylcholine onto muscarinic receptors of the gland. Apocrine glands are not thermoregulatory.
The secretory coil makes a primary fluid close to plasma in osmolality, and the duct reabsorbs sodium and chloride (chloride through CFTR, see membrane transport) while staying relatively impermeable to water. Sweat leaves hypotonic, and saltier at high flow, when the duct has less time per millilitre. Heavy sweating therefore loses more water than salt: plasma volume falls and osmolality rises, driving ADH release and thirst (water, sodium and potassium; body fluid compartments). With heat acclimation, aldosterone lowers sweat sodium, and sweating starts earlier and rises more steeply with core temperature. Evaporating sweat takes up about 580 kcal per kg, or 2.43 kJ per g at skin temperature (2.26 kJ per g is the value at 100 degrees C).
Worked problem. The numbers are made up for practice, except the latent heat and specific heats. A healthy 70 kg adult cycles steadily in hot, sunny air. Assume a metabolic rate of 1,000 W, mechanical efficiency of 20 percent, dry heat gain of 50 W, no respiratory heat loss, 1 g of sweat = 1 mL, and uniform body temperature for storage.
- Heat production: M - W = 1,000 W - (0.20 x 1,000 W) = 800 W.
- Required evaporation for S = 0: E_req = 800 W + 50 W = 850 W.
- Evaporation: 850 J per s / 2,430 J per g = 0.350 g per s, x 3,600 s per h = 1,259 g per h, about 1.26 L per h. Check: 1,259 x 2,430 / 3,600 = 850 W.
- If only 70 percent evaporates and the rest drips, sweat rate = 1,259 / 0.70 = 1,799 g per h, about 1.80 L per h. Without drinking, body mass falls about 1.80 kg in an hour, 1.80 / 70 = 2.6 percent.
- Humid air caps evaporation at E_max = 700 W. Storage S = 850 - 700 = 150 W, or 150 J per s x 3,600 s = 540 kJ per h. Heat capacity = 70 kg x 2.98 kJ per kg per degree C = 208.6 kJ per degree C, so temperature rises 540 / 208.6 = 2.6 degrees C per h. With the older, assumed 3.47 kJ per kg per degree C, 540 / 242.9 = 2.2 degrees C per h; the tissue-based value predicts heating about 16 percent faster.
- Once evaporation reaches E_max, sweating harder adds dripping, not cooling. Balance returns only with less heat production (slowing down), less dry gain (shade, cooler air) or more evaporation (drier or faster-moving air).
The rate holds only while heat production, dry gain and E_max stay fixed and body temperature is treated as uniform; it predicts no real person.
- Required evaporation E_req
- 850 W
- Sweat to evaporate to meet E_req
- 1,259 g per h
- Evaporation E (smaller of E_req and E_max)
- 700 W
- Heat storage S
- +150 W
- Rate of rise of body temperature
- 2.6 °C per h
- State
- uncompensable
In this sketch every number is illustrative and made up for practice, except the latent heat of evaporating sweat (2.43 kJ per g) and the two specific heats. Fixed, as in the lesson’s worked problem: a 70 kg adult, no respiratory heat loss, 1 g of sweat = 1 mL, uniform body temperature for storage, skin blood flow unchanged. Invented for the sketch: sweat is made at the rate that would meet E_req; any sweat beyond what E_max can evaporate drips and removes almost no heat (counted as none here); that dripping share is an assumption, not a measurement. The rate holds only while the three sliders stay fixed; it predicts no real person.
Rule to keep: evaporation is the smaller of E_req and E_max, and storage is what is left, S = (M − W) + dry exchange − E. The bars share one scale, 0 to 1,200 W.
The wet-bulb temperature, read from a wetted thermometer cooled by evaporation, combines air temperature and humidity. A wet-bulb of 35 degrees C was long cited as the theoretical limit, since fully wet skin near 35 degrees C could then lose no net heat to the air. Vecellio and colleagues (2022), in Kenney’s laboratory, raised heat stress in steps in young healthy adults at low metabolic rates. No one reached 35 degrees C: in humid air, core temperature stopped holding steady at a mean wet-bulb of about 30.5 degrees C, and in hot, dry air dry heat gain outpaced evaporation. Meade and colleagues (2025) then held 12 volunteers for up to 9 hours above and below each person’s inflection; above it, core temperature kept rising, and below it the rise was much slower but still did not level off. Both studied young adults at low activity in an indoor chamber.
Pathophysiology: exercise in hot, humid air. As air vapour pressure approaches that of wet skin, E_max falls below E_req and core and skin temperatures rise. Skin blood flow and heart rate rise. Cardiac filling and stroke volume fall, as sweat loss shrinks plasma volume and the faster heart shortens filling time; how much pooling of blood in skin veins adds during exercise is debated. Sweat keeps flowing, so plasma volume falls and osmolality rises. When dehydration adds to hyperthermia, the skin circulation joins the systemic vasoconstriction: in 7 trained cyclists in 35 degrees C heat, cutaneous vascular resistance rose despite a higher oesophageal temperature (González-Alonso and colleagues, 1995), cutting heat delivery to the skin. Their 1997 data are in exercise physiology. In mechanism, heat exhaustion is heat strain with weakness, faintness or a fast heart rate, without the marked central nervous system dysfunction of heat stroke, and heat stroke is severe heat illness defined by central nervous system dysfunction, such as confusion, seizures or coma, often with injury to other organs. Sweating often persists early in exertional heat stroke, so sweating does not mean heat balance.
In the cold: constrict, shiver, burn brown fat
In the cold, vasoconstriction limits loss and heat production rises, mainly from shivering muscle in adults (skeletal muscle mechanics); if core temperature keeps falling, shivering wanes below about 31 degrees C. Brown adipose tissue makes heat without contraction: UCP1 lets protons leak back across the inner mitochondrial membrane without making ATP (the proton-motive force; thyroid hormone).
Brown fat was thought to regress after infancy. In 2009 Wouter van Marken Lichtenbelt and colleagues scanned 24 healthy men, 10 lean and 14 overweight or obese, by 18F-fluorodeoxyglucose PET-CT after mild cold at 16 degrees C, with indirect calorimetry at 22 and 16 degrees C. After cold, 23 of 24 showed active brown fat, mainly in supraclavicular and neck depots, with lower activity in the overweight and obese men. Three men with high uptake, scanned again at 22 degrees C, showed none. In the same issue, Cypess and colleagues reviewed clinical scans retrospectively and found brown fat in 7.5 percent of women and 3.1 percent of men; those scans used no controlled cold exposure, and detection was more likely when outdoor temperature was lower.
The van Marken Lichtenbelt study does not show how much heat brown fat adds (FDG reports glucose uptake; brown fat burns mostly fatty acids), how women’s brown fat responds to cold (all imaged subjects were men), or whether low activity causes obesity or follows it.
Sunlight makes vitamin D in the living epidermis
UVB at 290 to 315 nm photolyses 7-dehydrocholesterol to previtamin D3, and a temperature-dependent isomerization, not light, converts that to vitamin D3. Most synthesis is in the living basale and spinosum, beneath the dead corneum. MacLaughlin, Anderson and Holick (1982) put the optimum at 295 to 300 nm.
Webb, Kline and Holick (1988) exposed excised human skin and radiolabelled 7-dehydrocholesterol to cloudless-day sunlight. In Boston (42.2 degrees N) no previtamin D3 formed from November through February, in Edmonton (52 degrees N) none from October through March, and at 34 and 18 degrees N it formed even in midwinter. Low winter sun sends light through more atmosphere, where ozone absorbs UVB. This is photochemistry in vitro, not blood levels in people. Melanin also absorbs UVB, so more pigmented skin makes less previtamin D3 per exposure. Sunlight has not been reported to cause vitamin D toxicity, because further UVB converts excess previtamin D3 to lumisterol and tachysterol, which do not act as vitamin D3. Activation in liver and kidney is in calcium, phosphate and bone.
Design problem
Do we lose most of our body heat through the head? Design an experiment in healthy volunteers in cold water that measures heat loss from the head and the rest of the body: null prediction, comparisons, measurements.
Saved in this browser only.
One way to do it
Null prediction: each region loses heat roughly in proportion to its surface area. The head is about 7 percent of the surface, so submerging it should add roughly that, not the myth’s 40 to 50 percent (the 1970 manual gave 40 to 45).
Use a 2 x 2 design: head in or out of the water, body insulated or not. Suppress shivering with a drug, so changing heat production cannot mask loss. Measure heat loss and the core cooling rate until both are steady.
Pretorius, Bristow, Steinman and Giesbrecht (2006) did this with 8 men in 17 degrees C water. With the body uninsulated, submerging the head added only about 10 percent to total heat loss, so the head did not lose disproportionately more. Yet the core cooled about 42 percent faster, possibly through redistribution of blood flow.
Limits: 8 men, one water temperature, no shivering, water not air. With a clothed body and bare head, the head is the main uninsulated surface and can carry a larger share. A 1970 US Army survival manual gave 40 to 45 percent; the experiment behind the figure has not been identified.
How we measure it
Electron-dense tracers and freeze-fracture electron microscopy
Introduce water-soluble tracers that show up in the electron microscope, such as horseradish peroxidase or lanthanum, and see how far they travel between cells; fracture frozen tissue along its membranes to view junctions and lipid. Together they map where a barrier lies.
18F-FDG PET-CT with controlled cold exposure
Inject a radioactive glucose analogue and image where it accumulates, with CT for anatomy, once at a thermoneutral temperature and once after mild cold. Uptake that appears only in the cold marks active brown fat; it reports glucose uptake, not heat output.
Environmental chamber step protocols
Hold a volunteer at a low, fixed metabolic rate while air temperature or humidity rises in steps, and record core temperature continuously. The point where core temperature stops holding steady and climbs estimates the limit of compensable heat stress for that protocol; longer exposures can still fail below it.
In vitro photoconversion in excised skin
Expose samples of human skin, or a radiolabelled precursor, to sunlight or to narrow bands of ultraviolet light, then measure how much 7-dehydrocholesterol became previtamin D3. It isolates the photochemistry from diet, clothing and blood levels.
Finish line
Test yourself
8 questions
Practice and MCAT-style, with explained answers.
Next · Lesson 27
Sexual Differentiation and the Testis
An AMH-null male mouse can make sperm and retain a uterus. Which testicular signals have separated?
Continue →