Story
What your body is doing during a fever
Last updated
In one pass A fever is not body temperature running out of control.
Educational content, not medical advice. See a clinician about your own case.
Story path
Chapter 1
The thermostat gets turned up
A fever is not body temperature running out of control. It is the brain deliberately raising its temperature target. That target is called the set point, and it is managed by a small area at the front of the hypothalamus called the preoptic area.
During an infection, innate immunity raises the first alarm: immune cells release a group of signalling molecules that cause fever, called pyrogens. They travel in the blood to the brain and act on the lining of its blood vessels, whose cells then make prostaglandin E2 (PGE2) on the spot. PGE2 acts on the preoptic area and releases its brake on the heat-making circuits; the body starts holding in heat and making more of it, and temperature is pushed toward the new, higher target.
Normal body temperature is not a fixed 37.0°C either. It is a range that shifts with where it is measured and with age.
The American Academy of Pediatrics stresses that fever is not an illness in itself but a physiological mechanism the body uses against infection. A baby under 3 months with a fever, or a child who cannot be woken or whose seizure will not stop, needs medical care right away.
During an infection, innate immunity raises the first alarm: immune cells release a group of signalling molecules that cause fever, called pyrogens. They travel in the blood to the brain and act on the lining of its blood vessels, whose cells then make prostaglandin E2 (PGE2) on the spot. PGE2 acts on the preoptic area and releases its brake on the heat-making circuits; the body starts holding in heat and making more of it, and temperature is pushed toward the new, higher target.
Normal body temperature is not a fixed 37.0°C either. It is a range that shifts with where it is measured and with age.
The American Academy of Pediatrics stresses that fever is not an illness in itself but a physiological mechanism the body uses against infection. A baby under 3 months with a fever, or a child who cannot be woken or whose seizure will not stop, needs medical care right away.
Mechanism · How the signal reaches the brain
The pyrogens in the main text include the interleukins IL-1 and and tumour necrosis factor (TNF). The key evidence for this step comes from mice: mice lacking IL-6, or with IL-6 neutralized by antibodies, do not develop a fever after being injected with the bacterial component lipopolysaccharide (LPS). This shows that in mice IL-6 is a required link in inducing fever.The route in the main text lands on two kinds of cell: cytokines act on the endothelial cells of brain blood vessels, and PGE2 binds EP3 receptors on preoptic neurons, releasing their inhibition of sympathetic neurons in the brainstem. Beyond this humoral route, researchers have proposed others, such as through the circumventricular organs, where the blood-brain barrier is weaker, through peripheral nerves, or by carrying PGE2 made elsewhere in the body into the brain; the review judges that the evidence for these is far from conclusive. Most of the data come from rodents.
Numbers · Normal temperature is a range
A 2019 systematic review pooled measurements from 36 studies covering 7636 healthy adults. Taking the mean plus or minus two standard deviations, rectal temperature ran roughly 36.32-37.76°C, oral 35.73-37.41°C and axillary 35.01-36.93°C; in the same person, a different measuring site gives a different reading. People aged 60 and over were about 0.23°C lower on average, and the difference between men and women was not significant. The authors concluded that the two most important personal factors in a reading are where it is taken and the person's age.These ranges are statistical reference intervals for healthy adults, not cut-offs for deciding whether someone has a fever. The review included only healthy adults, with no children and no one with a fever, and its search was limited to PubMed titles and English-language papers.
Chapter 2
Chills first, cooling later
Once the set point is raised, the brain judges the current body temperature to be too cold. So the body does what it would do standing in a cold wind: sympathetic nerves constrict the blood vessels in the skin, less blood reaches the skin, heat cannot escape, and the hands and feet feel cold to the touch; the muscles start to shiver, making heat by contracting. That is why a person can have a fever and still feel cold and want another blanket.
Once body temperature catches up with the new set point, heat production and heat loss are balanced again, and temperature holds at the higher level, still regulated by the brain.
When the infection signal fades and the set point comes back down, everything runs the other way: the current temperature now reads as too hot, the skin's blood vessels widen, and a large flow of blood goes to the skin to carry the heat away. The order is that the set point falls first and heat loss follows; losing heat is the result of the fever breaking, not the cause.
Once body temperature catches up with the new set point, heat production and heat loss are balanced again, and temperature holds at the higher level, still regulated by the brain.
When the infection signal fades and the set point comes back down, everything runs the other way: the current temperature now reads as too hot, the skin's blood vessels widen, and a large flow of blood goes to the skin to carry the heat away. The order is that the set point falls first and heat loss follows; losing heat is the result of the fever breaking, not the cause.
Mechanism · The wiring behind chills and cold hands
After PGE2 releases the brake in the preoptic area, the command travels down two routes. One runs directly from the preoptic area to the raphe region at the front of the medulla and drives constriction of the skin's blood vessels; the other is relayed first through the dorsomedial hypothalamus and then drives shivering and heat production by brown adipose tissue. This wiring was mapped mainly in rodents.Narrowing the skin's blood vessels uses the general equipment of temperature control: skin blood flow is controlled by sympathetic nerves, including a noradrenergic vasoconstrictor system that can push skin blood flow down very low. The review covers temperature regulation in general and did not study fever specifically; using it to explain cold hands and feet as a fever rises is applying the same mechanism to fever.
There is matching equipment for the other direction: the skin also has an active vasodilator system, and when body temperature is high, skin blood flow can rise to 6-8 litres per minute.
Myth · Sweating a fever out
The idea behind bundling up is that once you sweat, the fever breaks. By the set-point mechanism, cause and effect are the wrong way round. Before the set point drops, the body is already actively holding in heat, and extra blankets only make that heat harder to lose; once the set point comes down, the body opens the skin's blood vessels to shed heat on its own.The guideline on fever in children under 5 from the UK's National Institute for Health and Care Excellence (NICE) states two things: do not over-wrap a feverish child, and do not underdress them either; and tepid sponging to cool them down is not recommended.
Chapter 3
How fever reducers pull the set point back
The two common fever reducers, paracetamol (called acetaminophen in the US) and ibuprofen, both act on cyclooxygenase (COX), the enzyme that turns arachidonic acid into prostaglandins. With less prostaglandin being made, PGE2, the signal that raised the set point, weakens; the set point falls back, and the body then starts to shed heat.
The two drugs do not work in quite the same way. Ibuprofen is a non-steroidal anti-inflammatory drug (). Paracetamol instead works through COX's own peroxidase function: it is effective where arachidonic acid and peroxides are both scarce, and almost ineffective where both are plentiful. That is why it lowers fever and eases pain but cannot hold down severe inflammation such as rheumatoid arthritis.
Fever reducers improve comfort; they do not treat the infection itself. Paediatric sources set the goal of fever treatment as making the child comfortable, not bringing the temperature down to a particular number.
The two drugs do not work in quite the same way. Ibuprofen is a non-steroidal anti-inflammatory drug (). Paracetamol instead works through COX's own peroxidase function: it is effective where arachidonic acid and peroxides are both scarce, and almost ineffective where both are plentiful. That is why it lowers fever and eases pain but cannot hold down severe inflammation such as rheumatoid arthritis.
Fever reducers improve comfort; they do not treat the infection itself. Paediatric sources set the goal of fever treatment as making the child comfortable, not bringing the temperature down to a particular number.
Mechanism · When it works and when it does not
To make prostaglandins, COX first has to use its own peroxidase function to generate a key free-radical intermediate. Paracetamol is metabolized by exactly that peroxidase function, which stops the intermediate from forming, so the enzyme cannot make prostaglandins. This only holds when arachidonic acid and peroxide levels are low; when both are high, it has almost no effect.This matches what is seen in the clinic: it cannot hold down the severe inflammation of rheumatoid arthritis or acute gout, but it still dampens milder inflammation, such as after a tooth extraction. Two further points are separate observations: it behaves like an inhibitor that favours COX-2, with weak anti-platelet action and good tolerance in the gut; the review puts both down to its apparent COX-2 selectivity and does not say they are caused by the low-peroxide condition.
On average it is a weaker painkiller than but better tolerated. The review also notes that its pain relief involves central serotonin, opioid and cannabinoid systems, and that how it works was unclear for a long time.
Evidence · Two drugs, little difference
A 2011 clinical report from the American Academy of Pediatrics (AAP) found no substantial difference in safety or effectiveness between paracetamol and ibuprofen in generally healthy children. Using the two together lowers fever more than either alone, but combined use is more complicated and can lead to unsafe use.The NICE guideline for children is more specific: consider one of the two only when a child appears distressed, do not use them with the sole aim of lowering the temperature, and do not give both at the same time.
Evidence · Do fever reducers make fever last longer?
A common worry is that pushing a fever down might make the illness drag on. A 2013 systematic review found 6 studies in children, 5 of which went into the pooled analysis: in the groups given fever reducers, fever actually cleared about 4.16 hours faster on average (95% 1.96-6.35 hours), and statistical heterogeneity between the studies was low.Its limits need stating: there were few studies; the populations were mixed, with 3 in children with malaria and the other 3 in general viral or respiratory infections and chickenpox; and the outcome was only the time for the fever to clear, not the length of the whole illness or how fast children recovered. The authors concluded that these studies give no evidence that fever reducers slow the clearing of fever.
Chapter 4
Does fever help the immune system?
The higher temperature of a fever is not just a side effect; it is a signal in its own right. At fever-range temperatures, many immune cells change how they behave: innate immune cells such as macrophages, neutrophils and natural killer cells are stirred into action, and lymphocytes cross blood vessel walls into the lymph nodes more readily to patrol them. Fever exists in both cold-blooded and warm-blooded vertebrates and has been kept through a long stretch of evolution.
But this evidence comes mainly from whole-body heating of mice and from cell experiments outside the body. It shows a mechanism, not a treatment benefit in people; and once temperature goes beyond the fever range, neutrophil function is actually impaired.
What happens in people when a fever is pushed down? In a randomized trial in intensive care units (ICUs), feverish patients with suspected infection given a fever reducer had outcomes no different from those given a placebo. That conclusion applies only to critically ill patients: in them, pushing a fever down showed neither harm nor benefit. In children, the studies support only one point: fever reducers did not make fever clear more slowly.
But this evidence comes mainly from whole-body heating of mice and from cell experiments outside the body. It shows a mechanism, not a treatment benefit in people; and once temperature goes beyond the fever range, neutrophil function is actually impaired.
What happens in people when a fever is pushed down? In a randomized trial in intensive care units (ICUs), feverish patients with suspected infection given a fever reducer had outcomes no different from those given a placebo. That conclusion applies only to critically ill patients: in them, pushing a fever down showed neither harm nor benefit. In children, the studies support only one point: fever reducers did not make fever clear more slowly.
Evidence · Observational data from sepsis
One kind of human data is often cited to argue that fever helps. A 2016 review pooled retrospective ICU data: patients with sepsis whose temperature was raised on their first day in the ICU did better than those with normal temperature, with a trend toward improvement in the 37.5-39.4°C band. The authors' wording is that a mild rise in temperature may be associated with a survival benefit.This association has three limits. First, it is observational and cannot be read as fever causing survival. Second, above about 40°C, mortality rises again. Third, fever from non-infectious causes shows no such benefit, and in patients without sepsis, harmful effects start building up early, even with a mild fever.
Evidence · The HEAT trial
The most direct way to test whether pushing a fever down does harm is a randomized trial. The HEAT trial enrolled 700 ICU patients with a raised temperature and known or suspected infection, and randomly gave them intravenous paracetamol or a placebo.By day 28, the median number of days alive outside the ICU was 23 in one group and 22 in the other; 90-day mortality was 15.9% versus 16.6%; neither difference was significant. The authors concluded that giving paracetamol early to treat fever due to infection did not change the number of ICU-free days.
This is a null result: it shows neither that bringing fever down is harmful nor that it helps, and the patients were critically ill, so it cannot be carried over directly to ordinary people with a fever at home.
Chapter 5
Reading the thermometer and when to get help now
The number on a thermometer depends on where it is taken. Readings from peripheral sites such as the mouth, armpit, ear and forehead already differ from the body's core temperature, and the gap is wider during a fever; a normal reading cannot fully rule a fever out.
More important than the number is how the child looks overall. What follows is part of the red column of the traffic light grading that the UK's National Institute for Health and Care Excellence (NICE) uses for children under 5; it is not a complete list, and it does not apply to older children or adults. Skin, lips or tongue that are pale, mottled, ashen or blue; no response to social cues; not waking, or not staying awake once woken; a weak, high-pitched or continuous cry; a bulging soft spot on the head: NICE lists these as high-risk signs of serious illness, and any fever in a baby under 3 months also counts as high risk. In any of these situations, get medical help immediately.
Heatstroke and fever both mean a high temperature, but the mechanisms differ: in a fever the set point is raised; in heatstroke heat production outruns heat loss, core temperature climbs above the set point, and temperature regulation fails. Following the mechanism through: since the set point has not been raised, fever reducers have nothing to act on. Heatstroke is a life-threatening emergency: call emergency services and start cooling the body right away; how to cool someone is covered in the story on exercising in the heat.
More important than the number is how the child looks overall. What follows is part of the red column of the traffic light grading that the UK's National Institute for Health and Care Excellence (NICE) uses for children under 5; it is not a complete list, and it does not apply to older children or adults. Skin, lips or tongue that are pale, mottled, ashen or blue; no response to social cues; not waking, or not staying awake once woken; a weak, high-pitched or continuous cry; a bulging soft spot on the head: NICE lists these as high-risk signs of serious illness, and any fever in a baby under 3 months also counts as high risk. In any of these situations, get medical help immediately.
Heatstroke and fever both mean a high temperature, but the mechanisms differ: in a fever the set point is raised; in heatstroke heat production outruns heat loss, core temperature climbs above the set point, and temperature regulation fails. Following the mechanism through: since the set point has not been raised, fever reducers have nothing to act on. Heatstroke is a life-threatening emergency: call emergency services and start cooling the body right away; how to cool someone is covered in the story on exercising in the heat.
Numbers · How far off peripheral thermometers are
A 2015 systematic review pooled 75 studies covering 8682 patients, comparing peripheral thermometers (eardrum, temporal artery, armpit, mouth) with central measurement (pulmonary artery catheter, bladder, oesophagus, rectum). The researchers had set ±0.5°C as the acceptable error in advance, and the pooled 95% limits of agreement fell outside it.The gap is wider during a fever: in adults with a fever, the limits of agreement were about -1.44 to 1.46°C, meaning that at the same moment a peripheral reading could be more than a degree above or below core temperature. Peripheral thermometers detected fever with a sensitivity of 64% and a specificity of 96%: when one shows a fever, there usually is one, but it misses about a third of fevers.
Limits: most studies had a high or unclear risk of bias in how patients were chosen or how the test was done, the studies differed widely, and the rectum counted as a central reference here.
Red flag · Young babies and very ill children
Items in the NICE red column not listed in the main text include appearing ill to a health professional, grunting, breathing faster than 60 breaths a minute, moderate or severe chest indrawing and reduced skin turgor. Babies under 3 months with a temperature of 38°C or higher are high risk; the guideline also notes that some vaccines can themselves cause fever at this age.A 2021 guideline from the American Academy of Pediatrics deals specifically with term babies aged 8-60 days who look well and have a temperature of 38.0°C or higher, sorting them into three age groups for a systematic evaluation by a doctor. It covers only babies who look well; babies who look unwell and babies born preterm are outside its scope. For a baby who looks unwell, go by the NICE red column above.
pantell-2021-febrile-infants
Background · Febrile seizures
Some children have convulsions during a fever; these are called febrile seizures. The American Academy of Pediatrics 2008 guideline defines the simple type as a generalized seizure lasting less than 15 minutes, occurring once in 24 hours, in a child with no brain infection, no metabolic disturbance and no history of seizures without fever. About 2%-5% of children aged 6-60 months have one.A 2021 Cochrane review pooled 32 trials covering 4431 children: about 30% of children have another seizure, and giving fever reducers during a fever (paracetamol, ibuprofen, diclofenac) showed no benefit for preventing a recurrence, with moderate to high certainty of evidence. The NICE guideline likewise states that fever reducers do not prevent febrile seizures.
The review authors consider recurrent simple febrile seizures benign in nature, and argue that families should be given contact details for medical services along with information on recurrence and first aid.
The word benign here applies only to recurrences already confirmed as the simple type. By the definition above, a seizure lasting 15 minutes or longer is not simple. A child's first seizure, or one that will not stop or does not look like the description above, falls outside this conclusion: get medical help immediately.
aap-2008-febrile-seizureoffringa-2021-febrile-seizure-prophylaxis
References · 12
- Blomqvist, A., & Engblom, D. (2018). Neural mechanisms of inflammation-induced fever. The Neuroscientist, 24(4), 381-399. 10.1177/1073858418760481
- Evans, S. S., Repasky, E. A., & Fisher, D. T. (2015). Fever and the thermal regulation of immunity: the immune system feels the heat. Nature Reviews Immunology, 15(6), 335-349. 10.1038/nri3843
- Geneva, I. I., Cuzzo, B., Fazili, T., & Javaid, W. (2019). Normal body temperature: A systematic review. Open Forum Infectious Diseases, 6(4), ofz032. 10.1093/ofid/ofz032
- Sullivan, J. E., & Farrar, H. C.; American Academy of Pediatrics Section on Clinical Pharmacology and Therapeutics and Committee on Drugs. (2011). Fever and antipyretic use in children. Pediatrics, 127(3), 580-587. 10.1542/peds.2010-3852
- National Institute for Health and Care Excellence. (2019, updated 2021). Fever in under 5s: assessment and initial management (NG143). www.nice.org.uk/guidance/ng143/chapter/Recommendations
- Charkoudian, N. (2003). Skin blood flow in adult human thermoregulation: how it works, when it does not, and why. Mayo Clinic Proceedings, 78(5), 603-612. 10.4065/78.5.603
- Graham, G. G., Davies, M. J., Day, R. O., Mohamudally, A., & Scott, K. F. (2013). The modern pharmacology of paracetamol: Therapeutic actions, mechanism of action, metabolism, toxicity and recent pharmacological findings. Inflammopharmacology, 21(3), 201-232. 10.1007/s10787-013-0172-x
- Young, P., Saxena, M., Bellomo, R., Freebairn, R., Hammond, N., van Haren, F., et al. (2015). Acetaminophen for fever in critically ill patients with suspected infection. New England Journal of Medicine, 373(23), 2215-2224. 10.1056/NEJMoa1508375
- Purssell, E., & While, A. E. (2013). Does the use of antipyretics in children who have acute infections prolong febrile illness? A systematic review and meta-analysis. Journal of Pediatrics, 163(3), 822-827.e2. 10.1016/j.jpeds.2013.03.069
- Niven, D. J., Gaudet, J. E., Laupland, K. B., Mrklas, K. J., Roberts, D. J., & Stelfox, H. T. (2015). Accuracy of peripheral thermometers for estimating temperature: A systematic review and meta-analysis. Annals of Internal Medicine, 163(10), 768-777. 10.7326/M15-1150
- Walter, E. J., Hanna-Jumma, S., Carraretto, M., & Forni, L. (2016). The pathophysiological basis and consequences of fever. Critical Care, 20(1), 200. 10.1186/s13054-016-1375-5
- Epstein, Y., & Yanovich, R. (2019). Heatstroke. New England Journal of Medicine, 380(25), 2449-2459. 10.1056/NEJMra1810762
Written from the sources listed above. No physician reviews this site: it explains how the body works, and it is not medical advice. How we make content
Spotted an error, or something unclear? Email us. We read every message, check it against the sources, and fix what's wrong.