← Sage hub · Circadian / Energy Phase · Thyroid (tempo) · Surfer Temperature

Sage — Circadian Temperature

Boundary: This page is Temperature as a Circadian timing sub-pillar (thermal swing / 24h curve / Energy Phase weather). Related but distinct: Sage — Thyroid as signal/tempo governor (HPT axis). Cross-link; do not merge into a fifth pillar.

FRUGAL PHILOSOPHY ESSAY · SEPTEMBER 2026

A Metabolic Hypothesis of Temperature

By Val Zudans, MD

EnSurf · Temperature as the thermal chapter of chronic compensatory mode

Not medical advice. Informational purposes only. Physiology interpreted through a metabolic lens for metabolically curious adults, with qualified physician interpretation of any personal data. n=1 tracking is not a protocol.

The claim

Temperature is not a separate organ system. It is a running account of how a biology converts fuel into ATP, and of how much of that conversion is deliberately or unavoidably dissipated as heat.

We are born with perfect biology---perfect not for fluorescent office light, seed-oil emulsions, and air-conditioned nights at 72 °F, but for an ancestral pattern of fuel, light, load, and thermal swing.

The conventional story treats temperature as a thermostat problem: a hypothalamic set-point, some effectors, a few hormones. That story is not false. It is incomplete. It starts at the dashboard. The metabolic hypothesis starts at the engines.

When ATP-generating capacity matches load, temperature is easy. The organism can afford obligatory heat by day, a clean nocturnal decline, a defended fever when infected, and a quiet idle at night while mitophagy and repair run. When capacity chronically lags load, temperature becomes expensive. The system compensates---Warburg-like glycolysis, sympathetic drive, altered deiodination, inflammation, structural braces. The dashboard then fills with "disorders of thermoregulation," which medicine names and suppresses without asking why the engines were already running hot, or cold, or flat.

This essay gathers the temperature story under that single upstream question: what is the relationship between ATP-generating capacity and load, and what is heat doing in that relationship?

Heat is not optional

Food electrons move down the electron transport chain. Protons are pumped. A chemiosmotic reservoir forms across the inner mitochondrial membrane. ATP synthase collects a toll---classically on the order of three ATP per ten protons---and the remainder of biology is built on that toll-keeper.

The conversion is not perfectly coupled. It cannot be. Proton leak, redox slip, membrane cycling, and the Second Law guarantee that a fraction of substrate energy appears as heat. Endothermy did not invent heat. Endothermy organized heat: more leak, more futile cycles, more thyroid drive, brown adipose uncoupling when it exists, shivering when it must, and a brain that defends a narrow core.

Mitochondria themselves may run substantially warmer than the cytosol when the chain is active. The "hot mitochondrion" measurements remain contested on physics grounds---the apparent gradient is hard to reconcile with steady-state heat flow---but the biological point does not require a 50 °C matrix. The point is that the engines are furnaces. Heat is the inescapable companion of flux. A biology that treats heat as an afterthought has already misunderstood ATP.

EnSurf's targets follow from that: proper human diet matched to mitochondrial programming, circadian alignment (sleep, light, feeding windows, thermal therapies), ETC function, biogenesis, mitophagy, membrane integrity, and hormetic conditioning that resets apoptosis thresholds. Temperature sits inside every one of those targets. It is how you feel the engines from the outside.

The nested engines

Think of the body as a meta-engine generating power on several timescales.

OXPHOS is the base load. Fat-preferred. Slower. Higher yield. The nuclear reactor. It supplies the bulk of resting ATP and most of obligatory thermogenesis.

Cytoplasmic glycolysis is the mid-range turbine. Glucose-preferred. Faster. Far fewer ATP per carbon, but carbon units and speed when load exceeds OXPHOS. The Warburg shift is this turbine being called because the reactor cannot meet demand---or because growth needs carbon more than it needs efficient ATP.

The creatine-phosphate system is the hybrid battery. Cells sit loaded with creatine. Work converts ATP to ADP; CrP instantly recharges ATP. Rest and surplus recharge the battery. Fastest timescale.

Nested controls sit on top of the engines:

- AMP and related signals ask for more engines (biogenesis) and better engines (mitophagy).

• mTOR estimates future load.

- Testosterone and growth signals authorize structure only when the engines report surplus.

• Cortisol mobilizes substrate on a medium timescale.

• Autonomic tone mobilizes on a short timescale.

- Thyroid sets tempo: how hard the engines are asked to spin, how much of the energy is captured as ATP versus dissipated as heat, how many Na+/K+ and Ca2+ cycles run in the membrane.

Inflammation is not a random fire. It is a signal that load has exceeded capacity for too long. Calcium deposition is not a beneficial brace that just happens to appear in old vessels. It is structural memory of prior bioenergetic failure---an emergency scaffold thrown up around a chronic deficit. Fibrosis is the same story in other tissues. Chronic compensatory mode is the pathology we then name disease.

Temperature lives in this stack. Obligatory heat is the tax on ATP turnover and membrane cycling. Facultative heat is recruited when obligatory heat plus conservation cannot defend the core. Sleep is a scheduled downshift of the whole stack. Fever is a scheduled upshift of the defended core, paid for by the same effectors. Malignant hyperthermia is a leak in one of the futile cycles that will not turn off. Reye syndrome is what happens when you poison the hepatic engines while they are already covering a viral load.

Same physics. Different time-constants and different dashboards.

Thyroid as tempo governor

Thyroid hormone is how a mammal sets metabolic RPM and obligatory heat.

T3 increases mitochondrial biogenesis and respiratory-chain components. It increases Na+/K+-ATPase and SERCA cycling---classic ATP-consuming futile cycles whose product is heat. It modulates proton leak. It is permissive for UCP1 and for the sympathetic activation of brown fat. Centrally, it talks to the hypothalamus about set-point, TRH, and outflow to vessels and thermogenic organs.

Cold raises peripheral T3 conversion (D2 in BAT) and sympathetic drive so facultative thermogenesis can compensate for a colder world. Energy deficit, illness, and starvation lower active T3 and raise rT3---type-1 allostasis, conservation, less obligatory heat, less ATP demand. Anticipated demand (cold season, pregnancy, training) can push the other way.

Hypothyroidism is cold intolerance because both obligatory thermogenesis and the capacity for facultative thermogenesis collapse. Mild deficit can be patched by sympathetic tone; severe deficit cannot. Hyperthyroidism is heat intolerance because obligatory heat is excessive and cooling itself costs ATP (cardiac output, sweating).

A single morning TSH is one point on a wave. TSH rises in the late afternoon and evening, peaks in the early night, and is then inhibited by sleep. Stay awake and the nocturnal surge keeps climbing; morning TSH after a sleepless night can look "high" without a new gland failure. Nighttime TSH is also less bioactive than daytime TSH, which is why free T4 does not spike with the surge. Low-normal free T3 with rising rT3 and a "normal" TSH is often a dashboard of energy conservation, not a reason to ignore the engines.

Thyroid is cause and readout. Replace T4 without restoring conversion, receptor action, circadian phase, and mitochondrial readiness, and residual coldness, fatigue, and flat temperature curves persist. The gland was never the whole governor.

How the axis talks: TRH, TSH, T4, T3, TBG, rT3

The temperature wave and the thyroid wave are the same conversation written in two dialects. To read labs without worshipping a single TSH, you have to know who is speaking to whom.

TRH --- the hypothalamic request. Thyrotropin-releasing hormone leaves paraventricular neurons when the brain wants more tempo: cold, low circulating thyroid hormone at the hypothalamus, certain circadian phases. TRH tells the pituitary both how much TSH to make and how to glycosylate it---which is why not all TSH molecules are equally bioactive. We almost never measure TRH in clinic. You infer it from context: cold exposure, starvation, sleep loss, pituitary disease.

TSH --- the pituitary work order. TSH tells the thyroid to take up iodine, make thyroglobulin, couple iodotyrosines, and secrete mostly T4 plus a little T3. The pituitary does not see "how warm the person feels." It sees circulating T4 that it converts locally to T3 via deiodinase type 2 (D2) inside the thyrotroph. That is the first interpretive trap. TSH is a T4-at-the-pituitary sensor with a circadian wrapper and a sleep brake. It is not a tissue-T3 meter and it is not a mitochondrial readiness meter. Low tissue T3 with adequate circulating T4 can leave TSH "normal" while the engines are in conservation.

T4 --- the circulating reservoir. The gland's main product is thyroxine: four iodines, long half-life (on the order of a week), a stable pool. Most T4 is cargo, not the active signal. Peripheral tissues decide what to do with it.

T3 --- the tempo signal. Triiodothyronine is the molecule that binds thyroid receptors and writes the metabolic RPM: mitochondrial biogenesis, ion-pump cycling, uncoupling permissiveness, heart rate, heat. Only a minority of circulating T3 is secreted by the gland. Most is made outside the thyroid by outer-ring deiodination of T4. Free T3 is the unbound fraction that can enter cells. It is the closest routine blood proxy for "is tempo actually on," and it is still only a proxy. Local D2 can make T3 inside BAT, brain, and pituitary that never shows as a big swing in the venous draw.

TBG --- the shipping protein. Thyroxine-binding globulin carries most circulating T4 (transthyretin and albumin carry the rest). Greater than 99% of T4 and T3 is bound. Only the free fraction is biologically available. When TBG rises---pregnancy, oral estrogen, some liver states---total T4 and total T3 rise so the free hormone can stay roughly constant. When TBG falls---androgens, nephrotic loss, severe illness---totals fall. That is why free T4 and free T3 exist as assays, and why totals still matter when you suspect a binding-protein story. Estrogen-driven TBG is not hyperthyroidism. A free-hormone assay that is fooled by binding changes is not a gland failure.

rT3 --- the off-ramp. Reverse T3 is T4 with the other ring deiodinated (inner ring, deiodinase type 3). It is not the active tempo signal. In ordinary physiology it is a clearance intermediate. In starvation, carbohydrate deprivation, and non-thyroidal illness, D1 in liver slows: less T4 becomes T3, and rT3 is cleared more slowly, so serum rT3 rises while T3 falls. That is type-1 allostasis written in iodines---conservation, less obligatory heat, less ATP demand. Conventional endocrine often treats rT3 as a test you should not order. The metabolic reading is narrower: rT3 is useful when you already suspect conservation (low free T3, illness, under-eating, overtraining) and you want to know whether T4 is being parked rather than activated. It is not a license to add T3 by recipe. It is a dashboard of conversion under load.

The three deiodinases are the local governors the blood test cannot fully see:

- D1 (liver, kidney): makes circulating T3 from T4 and clears rT3. Sick, fasting, and some drugs slow it. That is the low-T3 / high-rT3 pattern.

- D2 (brain, pituitary, BAT, muscle): local on-demand T3. Cold and sympathetic drive induce D2 in BAT so facultative heat can run even if serum T3 is not high. The pituitary uses D2 to decide TSH---hence TSH tracks T4 more faithfully than it tracks muscle or liver T3.

- D3 (brain, placenta, sick tissues): the inactivator. T4 to rT3, T3 to T2. Fetal and ill tissue use D3 to keep local tempo down.

Feedback in one circuit: high free hormone at hypothalamus and pituitary lowers TRH and TSH. Low free hormone raises them. Sleep inhibits the nocturnal TSH surge. Cold and the clock raise the request. Cytokines, cortisol, and low carbohydrate availability push D1 down and D3 up. hCG in early pregnancy stimulates the TSH receptor and can suppress TSH while raising output. The axis is not a single thermostat. It is a conversation among a request (TRH), a work order (TSH), a reservoir (T4), a tempo signal (T3), a shipping protein (TBG), and an off-ramp (rT3), edited in each organ by deiodinases.

When and how to interpret the draw. Not medical advice. Patterns, not prescriptions. Same clock time, same conditions, interpreted with last night's sleep and last week's load.

- Draw in the morning, fasted if you also want glucose and insulin, before biotin (which wrecks some immunoassays), and---if you take thyroid hormone---usually before that day's dose so you are not sampling a T3 spike or an acutely suppressed TSH.

- TSH alone is a screening work-order, not a complete sentence. Pair it with free T4 at minimum. Add free T3 when symptoms and temperature curve disagree with TSH. Add rT3 when you are asking about conservation versus gland failure. Add TPO and thyroglobulin antibodies when autoimmunity would change the story. Total T4/T3 when pregnancy, estrogen, or a binding-protein puzzle is on the table.

- High TSH + low free T4: primary gland failure until proven otherwise. The work order is loud because the reservoir is empty.

- High TSH + normal free T4: the textbook "subclinical" pattern. It may be early gland failure, recovery from illness, or a set-point the pituitary is defending. Read the temperature curve and symptoms before you treat a number.

- Low or normal TSH + low free T4: look central (pituitary/hypothalamus) or severe non-thyroidal illness---not garden-variety Hashimoto's.

- Low TSH + high free T4/T3: hyperthyroidism or over-replacement. Heat intolerance is the thermal dashboard.

- Normal TSH + low-normal free T3 + rising rT3: conservation. Engines throttled. This is the pattern the single-TSH religion misses. Fix load, carbohydrate availability, sleep, and inflammation before you assume the gland is the problem.

- Normal everything + flat cold temperature curve + hypothyroid symptoms: possible conversion, receptor, or mitochondrial-readiness problem the venous draw cannot see. The curve is still data.

- After a sleepless night, morning TSH can be higher because sleep never applied the brake. Do not start a lifelong prescription on one post-call draw.

- Pregnancy needs trimester-aware ranges: TBG and totals up, first-trimester TSH often down because of hCG. Do not use non-pregnant cutoffs.

- On T4 replacement, TSH and free T4 tell you about supply to the pituitary. They do not guarantee peripheral T3. On T3-containing replacement, TSH is easily suppressed for hours after a dose; timing the draw is part of the measurement.

Reference intervals are population bells, not individual set-points. A TSH of 3.8 can be "in range" and still be a loud work order for that person. A free T3 at the bottom of the interval with a falling basal temperature is not a personality. Conventional practice is right that isolated rT3 and isolated free T3 should not be used to invent disease in a well person. The metabolic practice is right that a conservation pattern is information about engines, not a moral failing and not always a prescription for more hormone.

Read the conversation. Do not frame a single messenger.

The 24-hour temperature curve is the honest signal

In an entrained human, core temperature is not a flat 98.6. It is a wave.

CBT begins falling one to two hours before habitual sleep. Sleep onset is most likely when that decline is steepest. Distal skin warms---hands and feet vasodilate---the distal-proximal gradient widens, and heat leaves the core. Distal heat loss predicts sleep-onset latency better than the absolute core number. Sleep itself then further lowers metabolic rate and CBT. Warm-sensitive preoptic circuitry couples skin warmth, core cooling, and NREM. Melatonin participates in opening distal beds. The clock, the thermal program, and the HPT axis are the same conversation.

This is the chicken and the egg, resolved without mysticism.

The biohacker slogan says a drop in body temperature is "necessary" for sleep, and therefore an evening sauna or hot bath should be used to trigger compensatory heat loss. The physiology is real. The causality is slightly inverted.

The clock initiates a decline in heat production and an opening of heat-loss pathways. Sleep is the low-load state that that decline belongs to. Metabolic rate falls. OXPHOS can cover housekeeping with less glycolytic compensation. Surplus relative to load is what pays for repair, mitophagy, proteostasis, and possibly glymphatic flow, whose rise during sleep is debated.^1,2 The thermometer drop is both a signal (preoptic circuits will not reliably release wakefulness until the core is allowed to cool) and a consequence of engines downshifting.

Forcing a colder room when capacity is already short spends ATP on defense---vasoconstriction, restlessness, fragmented sleep. Passive heating one to two hours before bed works when it amplifies the natural sequence: raise core briefly, then dump heat through already-opening distal beds so the subsequent fall is larger. Do it too close to lights-out and residual core heat can steal slow-wave sleep even if efficiency still looks better. The bath is a prosthetic for a program that was supposed to be running. It is not proof that "colder" is the primary good independent of engine state.

Ideal is not the lowest nocturnal number. Ideal is amplitude and phase:

• a daytime plateau high enough to do work

• decline starting one to two hours before target sleep

• nadir in the second half of the night

• pre-wake rise

A flattened curve---high night temperature, missing evening decline---is a dashboard of circadian breakage, inflammation, or engines that cannot both produce heat by day and turn it down at night. Depression and some hypothyroid and allostatic states show that pattern: higher nocturnal core, blunted nocturnal TSH surge. The wave flattened. The engines did not get permission to idle.

The sleeping thermal nest

There is a sweet spot. It is not a single thermostat number. It is the zone in which the least ATP is spent on thermoregulation.

Two neutrals exist.

Room air, for a typical adult under ordinary bedding, clusters in sleep research around 60--67 °F (15.5--19.5 °C), often near 65 °F / 18 °C. Heat fragments sleep and cuts REM and slow-wave more than mild cool. Older adults sometimes need a slightly warmer room because the regulatory range narrows.

The sleeping body is different. Metabolic rate is lower than quiet wake. Bedding insulation is high. Naked sleep thermoneutrality sits nearer 28--32 °C. With a duvet, the bed microclimate can hold skin-comfortable temperatures near 30--32 °C while the room is 16--20 °C. Head and face stay cooler and dump heat. Covered trunk stays in a nest. That is the conserved strategy: warmth-seeking, curling, covering---then core cooling through distal beds.

Practical rule from the engine frame: room cool enough that you do not need sweating and extra cardiac output to shed heat; bedding warm enough that you do not shiver or uncouple merely to stay warm. Adjust insulation and coverage before you wage war on the whole-house thermostat. Humidity in a mid range matters because evaporation is part of the heat-loss pathway.

Thermal therapies belong in circadian alignment---timed heat, then a cool room and a warm nest---not as a club to beat core temperature down.

Thermal hormesis: training the engines, not forcing the thermometer

Sauna, hot tub, cold plunge, and cold shower are not sleep hacks and they are not fever. They are conditioning---the EnSurf target already named: hormesis training that resets apoptosis thresholds, expands the range the engines can tolerate, and buys time for biogenesis and mitophagy. Heat and cold are two directions of the same question: can the meta-engine accept a brief, bounded mismatch and come back more capable?

The goal is not a trophy core-temperature number. The goal is a more competent thermal and proteomic response so that ordinary life spends less ATP defending 37 °C. Repeated mild heat upregulates heat shock proteins, improves vascular conductance, and acclimates the cardiovascular system the way repeated hills acclimate a cyclist. Repeated mild cold upregulates the sympathetic--BAT--UCP1 axis, teaches distal beds to open and close on command, and trains the organism to rewarm from the inside rather than from a blanket. Both are load. Both must be paid. Neither replaces proper fuel or sleep.

What heat hormesis is paying for. Passive heat is an exercise mimetic for people who already have a heart that can accept the bill. Skin and core warm. Cutaneous vessels open. Heart rate and cardiac output rise. Sweat costs water and electrolytes. If core rises enough, the proteome is threatened and heat shock proteins---especially HSP70 and HSP90---are induced. Those chaperones refold damaged proteins, escort irreparable ones to degradation, and protect mitochondria and muscle from later insult. Finnish cohort data (frequent traditional sauna, typically around 80 °C, several times per week) associate dose with lower cardiovascular and all-cause mortality. That epidemiology is not a license to cook anyone. It is a hint that repeated, recoverable heat load is a conditioning signal.

Does core temperature have to rise? For the heat-shock and immune programs that look most like exercise, yes---some core rise is the more honest stimulus. A University of Oregon head-to-head in healthy young adults compared protocols as people actually use them: hot-water immersion (about 40.5 °C, 45 min), traditional sauna (three 10-minute bouts at 80 °C with cool breaks), and far-infrared sauna (about 45--65 °C, 45 min). Hot water raised core about 1.1 °C. Traditional sauna raised it about 0.4 °C. Far-infrared, in that protocol, barely moved core at all. Cardiac output and, uniquely, post-heat IL-6 and later immune-cell shifts followed the core rise. Water wins on core because you cannot evaporate heat while immersed; air saunas let you sweat the load away. Infrared heats tissue by radiation at lower air temperature; it can feel warm without a systemic thermal debt.

So is dry sauna better than infrared, hot tub, or bath? Better for what. If the target is a documented core-driven hormetic and cardiovascular pulse, hot-water immersion is the strongest single bout and traditional Finnish sauna is the modality with the long-term human outcome data. Infrared is often better tolerated and may still help people who cannot face 80 °C air; it is a weaker core stimulus in head-to-head work and should not inherit the Finnish mortality literature by brand name. A hot bath one to two hours before bed remains the better tool when the goal is the sleep heat-dump already described---not because it is a superior longevity protocol, but because the subsequent decline is the sleep signal. Same physics, different chapter.

What cold hormesis is paying for. Cold is a sudden increase in heat loss. Skin receptors fire. The cold-shock response occupies the first half-minute to two minutes: involuntary gasp, hyperventilation, heart-rate and blood-pressure spike, a large norepinephrine (and often dopamine) surge. That first window is real physiology, not folklore. A brief cold shower that actually feels cold can deliver mood, alertness, and a habituation-to-discomfort signal without a meaningful drop in core temperature. One cold-shower trial found that thirty seconds reduced sick days about as much as ninety seconds once a genuine shock occurred---diminishing returns on that particular dashboard.

Deeper and longer immersion is a different floor of the stack. Water at 10--15 °C over most of the skin extracts heat far faster than a shower. Core can fall half a degree or more. Norepinephrine rises further. Brown adipose tissue is adrenergically activated; UCP1 uncouples; fuel is burned as heat rather than captured as ATP. Repeated sufficient dose---on the order of minutes per session, totaling roughly ten-plus minutes per week of cool immersion in the protocols that mapped adult BAT---can increase BAT activity. That is facultative thermogenesis being trained, the same uncoupling biology fever and cold-season thyroid conversion already use.

Is the first thirty seconds enough? Enough for the shock: catecholamines, the gasp you must breathe through, the daily practice of not fleeing the signal. Not enough for a reliable BAT/UCP1 remodeling program, and not required for every claimed benefit. You do not need to become hypothermic. You also should not confuse a heroic core crash with virtue. The engines are being asked to make heat. If capacity is already short, a long ice sit is another mismatch, not medicine.

Heat shock proteins and cold shock proteins. They are cousins, not mirrors. HSPs (HSP70, HSP90 and others) are chaperones induced when proteins threaten to misfold---heat, exercise, fever, oxidative stress. They are quality-control staff for the proteome and, downstream, for mitochondrial integrity. A useful pulse is brief. Chronic HSP activation is a stress state, not a lifestyle. Cold-inducible RNA-binding proteins---RBM3 and CIRP are the names that appear in the cold literature---act more on RNA stability and translation than on protein folding. In animals they look neuroprotective. In humans they are a frontier, not a protocol. Do not treat "I did cold, therefore I made cold shock proteins" as interchangeable with the much better-measured HSP response to heat and exercise. Two languages. One theme: a bounded threat induces maintenance programs.

Energy management: active day versus recovery night. Thermal hormesis is load. Place it where the engines can pay.

- Morning or early active phase favors cold if the goal is alertness and sympathetic tone. Catecholamines belong with daylight, movement, and work. A short plunge or cold finish to a shower is a starting gun, not a recovery tool.

- Evening favors heat timed one to two hours before sleep if the goal is the circadian heat-dump. Finish early enough that core is falling at lights-out. A late ice bath that leaves you wired is the wrong floor of the stack for night.

- After endurance work, heat is generally compatible with recovery blood flow and HSP-mediated protein repair.

- Immediately after heavy resistance training, cold immersion can blunt the hypertrophic signal (muscle protein synthesis, satellite cells). If the session's purpose was to tell muscle to grow, do not ice the message. Move cold hours later or to a non-lifting day.

- Contrast (heat then cold, often two or three rounds) is a vascular pump plus a large norepinephrine spike if you end cold. The Søberg-style metabolic reading of ending on cold is that the organism must rewarm internally---BAT and muscle thermogenesis---rather than borrowing heat from a shower. That is extra engine work. Use it when surplus exists.

- Do not stack thermal load on depleted reserve: illness, raw sleep debt, unstable heart disease, uncontrolled hypertension, pregnancy without guidance, or a child. Hormesis that cannot be repaid is just stress.

Shock and core change are both real and they are not the same invoice. The first thirty seconds of cold is an autonomic and attentional event. A 0.4--1.1 °C core rise from heat is a systemic metabolic and proteomic event. A measurable core drop from immersion is a thermogenic event that recruits uncoupling. Choose the invoice that matches the goal. None of them replace the wave: surplus by day, permission to idle by night. Thermal hormesis is how you practice defending that wave under a brief, voluntary mismatch---so that an involuntary mismatch later costs less.

Condition the engines. Do not worship the thermometer.

Fever: defense, command, then heat

Fever is the question that forces the hypothesis to be precise.

Is fever overheating engines? Uncoupling? A primitive defense? All of them, in a specific order.

First it is a command. Exogenous pyrogens (LPS, viral products) and tissue injury induce endogenous pyrogens---IL-1β, IL-6, TNF-α, interferons. Those signals, via COX-2 and PGE2 acting at EP3 receptors in the preoptic area, change the firing of warm-sensitive neurons. The defended core is raised. This is not a broken thermostat. It is a new set-point. The organism now defends 39 °C the way it previously defended 37 °C: you feel cold, you vasoconstrict, you seek blankets, you shiver, you recruit brown fat if you have it, you run muscle futile cycles.

Then it is effector physics. The new set-point is paid for with the same heat sources already discussed:

• cutaneous vasoconstriction (conserve)

• shivering (ATP hydrolysis in muscle, heat as byproduct of work)

- BAT UCP1 where present (true mitochondrial uncoupling: protons leak, flux rises, ATP yield falls, heat rises)

- sarcolipin--SERCA uncoupling in skeletal muscle (ATP hydrolyzed to pump calcium that immediately leaks back; heat without useful work)---a plausible maintenance mechanism in adult humans whose BAT is limited

- increased obligatory turnover as sympathetic and thyroid-permissive drives rise

So yes: fever uses uncoupling and futile cycles. It is not caused by engines spontaneously overheating. The brain asked for a hotter core. The engines were conscripted.

And it is a primitive defense. Many pathogens have thermal optima below febrile range. Innate immunity runs hotter chemistry. Iron is sequestered. The sick organism reduces foraging and social contact. Fever is costly---every extra degree is a large increment in energy expenditure---which is why it is reserved for genuine threat and why it is dangerous when engines cannot pay.

The combination is the truth. Defense is the purpose. Set-point change is the command. Uncoupling, shivering, and SERCA cycling are the payment. "Overheating engines" as a first cause is the wrong order except in the special cases below.

Suppressing fever is therefore not a free dashboard wipe. Sometimes the cost of the fever exceeds the benefit---extreme hyperthermia, limited reserve, certain host factors. Sometimes blocking COX removes the command and the organism stops paying for a defense it still needed. The metabolic reading is: ask whether the engines can afford the bill, and whether the bill is still pointed at a real threat. Do not treat the thermometer as the disease.

The Reye pitfall: do not poison the hepatic engines during a viral load

Reye syndrome is the canonical warning against dashboard medicine at the thermal interface.

It is a rare, acute, noninflammatory encephalopathy with microvesicular fatty liver, almost always in children or adolescents, classically after influenza or varicella, and strongly associated with salicylate use during that illness. Blood sugar falls. Ammonia and acidity rise. The brain swells. Mitochondria in liver are injured. Fatty-acid oxidation and ureagenesis fail. The picture resembles inborn errors of mitochondrial β-oxidation and the urea cycle---and those occult errors are part of the susceptibility.

Aspirin here is not acting primarily as an antipyretic success story. Salicylates, especially in overdose or in a stressed child, interfere with mitochondrial function: uncoupling and TCA effects at toxic doses, impaired handling of long-chain fatty acids, a second hit on engines already covering viral load. The same drug class that lowers PGE2 (and therefore can lower the fever command) can also damage the organelles that must pay every other bill the liver has.

That is the pitfall in one sentence: you treated the thermal dashboard with a mitochondrial toxin while the engines were already in compensatory mode.

High-dose salicylate poisoning in any age makes the physics even plainer: uncoupling, heat, lactic acidosis, hyperventilation, neuroglycopenia. Hyperthermia in that setting is engine-level uncoupling running away. It is not PGE2 set-point fever. Treating those two hyperthermias as the same object is how you get philosophy and children hurt.

The metabolic rule is ugly and simple. Fever is expensive but often purposeful. The antipyretic that also uncouples or blocks fatty-acid oxidation in a virally loaded pediatric liver is the wrong tool. Capacity was the constraint. The dashboard was not the disease.

Malignant hyperthermia: a futile cycle that will not stop

Malignant hyperthermia is relevant. It is not "fever." It is not infection. It is a pharmacogenetic rupture of muscle calcium handling that turns a normal thermogenic mechanism into a runaway bill.

Most cases trace to gain-of-function variants in RYR1 (sometimes CACNA1S, STAC3). The ryanodine receptor is the sarcoplasmic reticulum calcium-release channel. Volatile anesthetics and succinylcholine open a leaky channel. Calcium floods the myoplasm. Contraction machinery activates. SERCA tries to pump calcium back in---ATP hydrolysis without a useful steady state, because the leak continues. Oxygen consumption and CO2 production soar. Heat soars. Membranes fail. Potassium and CK spill. Acidosis. Death if untreated.

This is muscle-based nonshivering thermogenesis gone neoplastic. The same Ca2+ cycling that can contribute modest heat in ordinary mammals becomes, with a leaky RyR1, a furnace that will not accept the off command. Early theories that MH was primary mitochondrial uncoupling did not hold. The mitochondria are downstream victims and copayors: they are asked for impossible ATP flux to feed SERCA. The primary leak is the calcium channel.

Heat here is not a raised set-point. The hypothalamus did not ask for 42 °C. A peripheral cycle asked for infinite ATP and dumped the difference as heat. Dantrolene closes the channel. That is engine-adjacent repair, not antipyresis.

MH belongs in this essay because it proves the category distinction the hypothesis needs:

• Set-point fever --- brain commands a higher core; effectors obey.

- Uncoupling/futile-cycle hyperthermia --- a leak forces flux and heat regardless of set-point (salicylate toxicity, MH, some toxic uncouplers).

- Failure of cooling --- capacity or environment cannot dump heat (hot-wet bulb limits, anticholinergic anhidrosis, extreme exertion).

- Failure of production --- engines cannot make enough heat (severe hypothyroidism, starvation, some mitochondrial disease).

Four dashboards. One physics.

Laboratory metrics as flux instruments

Labs are not diseases. They are probes of capacity, load, substrate choice, compensation, and tempo.

Thyroid tempo. TSH with clock-time and last night's sleep attached. Free T4 as reservoir. Free T3 as active tempo. rT3 and T3/rT3 as conversion versus inactivation. Interpret the wave, not the isolated morning TSH.

Substrate and compensation. Fasting glucose and insulin / HOMA: is the glycolytic turbine being over-called at rest? Resting lactate: glycolytic overflow when OXPHOS cannot keep up. Ketones: fat-OXPHOS online. Triglycerides and HDL: chronic overflow and hepatic compensation. RQ if you ever measure it: 0.7 fat, 1.0 carb; a resting RQ stuck high is inflexibility.

Capacity and stress. RMR or sleeping metabolic rate, when available. GDF15 and FGF21 as circulating energy-resistance / mito-stress signals. CK and LDH as muscle leak. Ferritin and iron studies: cofactor versus sequestered load. CBC and lymphocyte-to-monocyte ratio: adaptive immunity is expensive; lymphocyte charge is a readout of whether ATP supply supports it. Grip and a VO2 proxy: volume and quality of engines you can actually use.

Load and recovery. Cortisol slope, not a lone morning draw. Overnight resting heart rate and HRV: did the stack downshift? hs-CRP: inflammation as mismatch flag. Sex hormones and IGF-1: anabolism authorized only in surplus.

A random 8 a.m. panel cannot show phase or amplitude. That is why the temperature curve and the TSH curve belong on the same dashboard as HOMA and grip.

An EnSurf metabolic dashboard for temperature

Cheap daily layer---the operating system:

- Body-temperature curve, not a single number. Daytime plateau, evening decline, nocturnal nadir, pre-wake rise. Morning basal temperature is a crude expression proxy. Shape is the flux story. Warm hands and feet before sleep are the DPG story.

- Sleep environment and timing. Light, last meal, cool room, warm nest. Timed heat if you use it, then the cooling window.

- Overnight autonomic. Resting HR down, HRV up, few thermal awakenings. Engines successfully idling.

Periodic lab layer, same conditions each time:

• TSH, fT4, fT3, rT3

• fasting glucose, insulin, lactate; optional ketones

• lipids, hs-CRP, ferritin/iron, CBC (LMR)

• optional cortisol slope, sex hormones, GDF15

The research-grade TSH curve would show a late-day rise, an early-night peak cut by sleep, low daytime values. A robust surge that sleep is allowed to suppress is healthier than a high flat line or a missing wave. You will almost never get that curve clinically. Treat morning TSH as one point on a wave.

The frugal dashboard already in use---HOMA-IR, HbA1c, hsCRP, LMR, TG/HDL, blood pressure and nocturnal dipping, AST, visceral fat and waist-to-height, lean mass, grip, a VO2 proxy, cortisol shape, overnight HR---should have the temperature wave sitting next to them. Temperature is not a wellness gadget. It is the cheapest continuous calorimeter you own.

What chronic compensation looks like in the thermal domain

If the hypothesis is right, disordered temperature is usually not a primary thermostat disease.

Low daytime temperature, cold intolerance, and a flattened nocturnal dip after years of mismatch: engines downregulated (low T3, low SNS, low muscle mass), or engines present but inefficient, or both. The scar story appears as the usual suspects---visceral fat as defended energy, vascular calcium as archive, fibrosis as brace, Warburg-like lactate as turbine compensation.

Heat intolerance with a "normal" thyroid panel: cooling is ATP-costly; autonomic effectors are clumsy; local conversion is off; or obligatory heat is high because leak and inflammation will not stand down.

Insomnia in a warm room is not a character flaw. The preoptic gate did not get its heat-loss signal. Insomnia in a person who cannot generate a decline because the engines never downshift is a circadian-plus-capacity problem, not a missing milligram of melatonin.

Fever that will not pay: frail reserve, mitochondrial disease, severe hypothyroidism, or a child whose liver engines just took aspirin on top of influenza. Fever that will not stop: ongoing pyrogen, or a leak (MH, toxin) that is not a set-point problem at all.

Fix the engines---ancestral-pattern fuel, circadian and thermal alignment, strength as a biogenesis signal, zone-2 as quality, hormesis that raises thresholds without chronic overload---and the thermostat has less work. Suppress the thermostat and leave the mismatch, and you have purchased a quieter dashboard.

Closing

We are born with perfect biology for a world of real food, real light, real load, and real thermal swing. Heat is the tax on being alive at 37 °C instead of at pond temperature. Thyroid writes the tax rate. The clock writes the payment schedule. Infection raises the assessed value for a few days. Sleep is the audit that lets repair happen when the bill is small.

The metabolic hypothesis of temperature is only the thermal chapter of a larger claim: aging and chronic disease are chronic compensatory mode in a biology whose ATP-generating capacity no longer matches load. Calcium, fibrosis, inflammation, Warburg chemistry, flat hormones, flat temperature waves---these are not separate mysteries. They are what a thwarted persistence looks like from different floors of the same building.

Do not force the thermometer. Restore the conditions under which a wave can exist: surplus by day, permission to idle by night, enough reserve to afford a fever when a fever is the honest response, brief voluntary heat and cold so the engines learn the range, and enough sense not to uncouple a child's liver while the wave is already a defense.

The engines were always the point.

EnSurf: mitochondrial optimization as the upstream lever. Proper human diet. Circadian alignment, including thermal. ETC, biogenesis, mitophagy, membrane, conditioning. Fix the engine. Let the dashboard follow.

Parent pillar: Sage — Circadian / Energy Phase. Related: Sage — Circadian Light. Tempo governor deep-dive: Sage — Thyroid. Coach: Coach — Circadian Temperature.

A working hypothesis, open to correction and revised as new evidence arrives. Last revised November 1, 2026.

References

1. Xie L, Kang H, Xu Q, et al., 2013. Sleep drives metabolite clearance from the adult brain. Science 342(6156):373-377. doi:10.1126/science.1241224
2. Miao A, Luo T, Hsieh B, et al., 2024. Brain clearance is reduced during sleep and anesthesia. Nature Neuroscience 27(6):1046-1050. doi:10.1038/s41593-024-01638-y

See the notice at the top of this page. Working hypothesis, not medical advice.