How to Control Your Cortisol & Overcome Burnout

Huberman opens the episode with an unusually strong claim for a science podcast: of everything in the archive, this is the installment he says listeners could most benefit from, on the grounds that cortisol touches mood, sleep, immune function and long-term wellbeing, and that its timing is 'one of the most, if not the most powerful levers for your health and well-being.' The rhetorical move matters, because the rest of the episode is built to justify it: the argument is not that cortisol is dangerous, but that its daily shape is controllable, and that most of the suffering attributed to 'high cortisol' is actually suffering from a badly placed curve.

The central correction is semantic and, he insists, practical. Popular usage pairs cortisol with stress so tightly that any rise reads as damage. He accepts the pairing has a basis, since cortisol is made and released in response to stress, but argues the framing hides the hormone's other jobs, which he calls 'super important and positive.' The distinction he wants listeners to hold is conditional rather than absolute: the consequences depend on how high the levels are and when, which is why the episode is organized around a rhythm instead of a list of dos and don'ts.

'Cortisol is not a stress hormone per se.' He repeats the line immediately for emphasis, then supplies the replacement definition: cortisol is involved in deploying and directing energy to the tissues that need it most, above all the brain. On this account the hormone is not a damage signal but a fuel-allocation signal, mobilized for stressors and for everything else that demands output, including getting out of bed, a hard exam, a difficult conversation at work, or driving children to school.

The physiology he supplies is deliberately concrete. Cortisol raises blood glucose by driving energy release from the liver and from muscle, and he claims it can even cause glucose release from individual neurons. Critically, cortisol is not made in the brain; it is made in the adrenal glands, the same glands that release epinephrine, also called adrenaline. That shared origin conceals a timing difference he returns to later: the adrenals can release epinephrine almost instantly, but cortisol must first be synthesized and then released, so it operates on a slower clock.

A mechanistic detail carries much of the later argument. Cortisol is lipophilic, and because cell membranes are made of lipid, the hormone can pass through them. Practically, this means cortisol released from the adrenals into the blood can cross the blood-brain barrier, unlike adrenaline, and the brain is well prepared for it: there is a dense population of cortisol receptors in the hippocampus, a structure he flags as essential for understanding chronic stress and notes he will return to. The claim being assembled is that cortisol is a brain-directed energy hormone, not merely a systemic alarm.

He states the takeaway twice in different registers, first as an aspiration that listeners and 'hopefully the rest of the world' will stop calling cortisol a stress hormone, then as a definition: cortisol's main job is to deploy energy, making it an energy-producing hormone and, in particular, a brain energy-producing hormone. Listeners should notice what this reframing buys him. If cortisol is fuel allocation, then a high morning level is not a symptom to suppress but a resource to secure, and the clinical question becomes scheduling rather than lowering.

Rather than list protocols, Huberman defers them until the rhythm is on the table, and the reason is explicitly strategic. Cortisol is not a static quantity that can be raised or lowered independent of clock time; it changes continuously across the 24-hour cycle. Without a model of how that endogenous rhythm is generated, he argues, a listener cannot know which tool to apply when, and the same intervention could push the curve in a helpful or unhelpful direction depending on the hour.

The evidence he leans on is a 'quite classic study' examining the 24-hour pattern of episodic cortisol secretion in normal subjects, which he says identified essentially four phases. He reports the finding without naming authors, journal or year, a gap worth noting: the four-phase description is presented as settled background, and the listener is asked to accept the periodization on his authority rather than to evaluate the primary source. The numbers themselves, however, are specific enough to be checkable against sleep timing.

Phase one: a six-hour stretch of 'minimal secretory activity,' beginning four hours before lights out and continuing until two hours after lights out, when cortisol should be very low and stay low. Phase two: a three-hour window, the third to fifth hour of sleep, which the study called 'the preliminary nocturnal secretory episode,' in which cortisol begins to rise slightly. Phase three: the main secretory phase, listed as four hours though he says the exact duration varies, occupying roughly the sixth through eighth hour of sleep, when cortisol climbs very fast. Phase four: the post-waking period, to which he adds his own sub-phase, the first hour after eyes open, described as a window of opportunity that does not recur until the next morning.

The practical consequence is that sleep duration determines how much of the pre-waking cortisol surge a person actually receives. He anticipates the objection directly: a six-hour sleeper does not get the fast rise at all, a seven-hour sleeper gets only a portion, an eight-hour sleeper gets the full duration, and a nine-hour sleeper keeps rising. This converts a seemingly abstract endocrine curve into a concrete cost of short sleep, and it does so without claiming that short sleepers are cortisol-deficient in any clinical sense, only that they miss a scheduled pulse.

He then supplies a functional rationale for why the surge lands where it does. Early sleep is dominated by slow-wave deep sleep, when growth hormone is released and metabolism runs low; low metabolic demand means low glucose requirement, which he ties to low cortisol. The final hours are when REM sleep is most abundant, and REM is described as a period of 'robust brain activity that even is greater than what occurs during wakeful states.' The inference is left for the listener to complete but he states it anyway: vigorous brain activity needs fuel, and cortisol is what deploys the fuel.

The shape he wants listeners to visualize is asymmetric. Cortisol should climb steeply just before and after waking: 'Steep, steep slope of increase.' The first hour after opening the eyes, whether by alarm or spontaneously, is when that rise can be amplified further, and he stresses the window closes after roughly 90 minutes. Then, some three to four hours later, levels should begin a gradual decline that continues through the afternoon and evening until the low point before the next pre-dawn ascent.

Both directions of error are flagged as harmful. A decline that is too slow is bad; a decline that is too sharp is also bad. What he wants is a sharp morning spike followed by a taper that is neither a cliff nor a plateau, arriving at low levels a few hours before sleep. And he closes the loop with a claim that reframes waking itself: people wake up because of a rise in cortisol, 'the cortisol awakening response, or CAR.' That response, he says, is the anchor point where control is available, and getting the first hour right sets up the nighttime pattern as well.

The mechanistic section begins at the adrenal gland, specifically a middle layer where the cells that manufacture cortisol reside. Those cells do not act on their own; they wait for an instruction in the form of a hormone, and the chain that delivers it starts in the hypothalamus, described as a structure about the size of two small marbles sitting above the roof of the mouth, adjacent to a ventricle where cerebrospinal fluid circulates.

The ventricular location is not incidental. Because blood vessels and capillaries line the ventricles, neurons there are exposed to more of the body's blood chemistry than neurons buried deeper in the brain, which gives them what amounts to a readout of circulating hormones. The neurons in question sit in the periventricular nucleus, and he pauses to unpack the name, since 'para' means near and 'ventricular' means ventricle. These cells extend processes into the pituitary, the two-lobed stalk he compares to two garlic cloves, one with a thicker stalk than the other.

The signaling cascade runs as follows: periventricular neurons release corticotropin-releasing hormone, or CRH, which acts on the anterior pituitary and causes release of ACTH, adrenocorticotropin hormone, into the bloodstream. ACTH travels to the adrenals and binds the melanocortin receptor in that middle layer, triggering synthesis and release of cortisol. He flags the receptor name as seemingly unimportant now but relevant later, because melanocortin is involved in pigmentation, which is how he explains the blotchy facial pigmentation seen in Cushing syndrome, the condition of chronically elevated cortisol.

What makes the system self-regulating is that the same periventricular neurons that issue the command can also read the result. As cortisol enters the blood, those cells sense how much is present, functioning, in his image, as a thermostat rather than a thermometer. When cortisol is low, they release more CRH, driving ACTH and then more cortisol; when cortisol accumulates, they shut the signal down. He notes that cortisol's relatively long half-life means it accumulates, which is precisely what makes the sensing possible.

'So this is what we call a negative feedback loop.' He points out that testosterone operates the same way and that estrogen is similar though subject to additional regulation. The functional payoff is that cortisol shuts down its own production once it crosses a threshold, which is why a morning peak is followed by a decline rather than indefinite elevation, and why a later demand for energy produces only a bump riding on an already falling rhythm rather than a fresh plateau.

He walks the loop through a full day to show how the low point is generated as actively as the high one. As levels fall through the afternoon and evening, the CRH cells register biochemically that cortisol is very low, not zero but low enough, and resume releasing CRH, which restarts ACTH and adrenal output. That is why cortisol rises again late in sleep. The summary is deliberately mechanical: cortisol regulates its own production so it does not get too high or too low, using hypothalamic cells as the sensor.

The whole arrangement is named the HPA axis, hypothalamic-pituitary-adrenal, and he assigns it two jobs: generating the 24-hour circadian rhythm and generating the cortisol response to stress, both short-term and long-term. This dual role sets up the episode's most useful timing distinction, because the same axis serves a slow rhythmic function and a fast reactive one, and the two interact.

The timing distinction is where the account becomes explanatory rather than descriptive. A stressor produces an almost immediate response, heart rate and blood pressure up, pupils and eyes widened, alertness and attention shifted. That speed comes from epinephrine and norepinephrine released from the adrenals and from the brain, acting within milliseconds if not seconds, as in the example of jumping out of the way of a car that veers onto the sidewalk. Cortisol is also deployed, but it must be synthesized, released, and then mobilize glucose.

'In fact, it takes about 10 minutes.' That single number explains a common experience he describes: the stressor resolves, the car misses you, and yet five or ten minutes later you are still shaken. Because cortisol has a much longer half-life than adrenaline or norepinephrine, the physiological wave arrives after the event and outlasts it. He adds a detail with mechanistic weight: adrenaline cannot cross the blood-brain barrier, so the brain releases norepinephrine in parallel from its own sites, since waiting for blood-borne adrenaline would be too slow for a reaction that must happen now.