Adrenaline released during a short burst of stress dispatches immune cells from the spleen, narrows attention onto the task in front of you, and speeds wound healing. Stanford neurobiologist Andrew Huberman argues that the popular picture of stress as an evolutionary leftover, a tiger alarm misfiring in an office chair, gets the biology backwards. The stress response was never built for tigers. It was built to mobilize the body against anything: cold, famine, infection, a partner who didn't come back from the hunt, a deadline. The damage comes from duration, and from never having learned how to press the brake.
On that framing, "managing stress" splits into three distinct timescales, each with its own physiology and its own set of tools: real-time breathing for the acute spike, deliberate mind-body dissociation for the weeks-long grind, and social connection plus selective supplementation for the chronic state that erodes cardiovascular health.
The timing of this argument matters. Breath work, cold exposure, and cortisol-lowering supplements have all gone mainstream without much explanation of mechanism, and the same people buying ice baths are also being told stress will shrink their hippocampus. Sorting those claims by timescale, and anchoring each tool in a named circuit, lets a reader tell which intervention fits which situation.
The stress response is a generic yes-and-no switch
Huberman separates stressors (the things that happen to you) from stress (your psychological and physiological response). The response itself doesn't distinguish between standing in the cold without a jacket and cramming for three exams while sleep-deprived. Both trigger the same machinery.
That machinery is the sympathetic chain ganglia, a column of neurons running roughly from the neck to the navel. When a stressor registers, the whole chain fires at once, "like a bunch of dominoes falling." Those neurons release acetylcholine, which in turn drives postganglionic neurons to release epinephrine (adrenaline) at target organs throughout the body.
Epinephrine then does two opposite things at once. At beta receptors on the heart and leg muscles, it dilates blood vessels and speeds the heart. At receptors on the digestive and reproductive systems, it constricts vessels and shuts those functions down. Huberman calls this two-pronged action the key to understanding stress: it's a yes for the systems that support movement and a no for the systems that are luxuries when things are going well. Dry throat during a presentation is simply blood flow being diverted away from the salivary glands.
The third feature follows from the first two. Because the entire system is designed to make you move, unspent activation shows up as agitation, tremor, or the urge to say something you'll regret. Any tool that works has to deal with that agitation directly, in the body, rather than through instruction.
Telling yourself to calm down, Huberman says, tends to make it worse. Telling someone else to calm down does the same.
The diaphragm is the one internal organ you can steer, and it drives heart rate
The fastest brake he knows of exploits an anatomical accident: the diaphragm is a skeletal muscle. Unlike the pancreas or spleen, it can be moved on command through the phrenic nerve, and its movement changes heart rate through a loop that appears in every medical textbook as respiratory sinus arrhythmia.
The mechanism runs like this. On an inhale the diaphragm drops, the chest cavity expands, and the heart gets slightly larger. Blood moves more slowly through that larger volume. The sinoatrial node registers the slowdown and signals the brain, which sends a command back to speed the heart up. On an exhale the reverse happens: the heart compresses, blood moves faster, and the parasympathetic branch, via neurons in the brainstem, slows the heart down.
The rule that falls out of this is simple enough to remember under pressure. Inhales longer or more vigorous than exhales raise heart rate. Exhales longer or more vigorous than inhales lower it.
The physiological sigh adds a twist. It's a double inhale through the nose followed by a long exhale through the mouth, a pattern humans and animals do spontaneously before sleep, after hard crying, and in cramped spaces. The second inhale, even if it sneaks in only a little extra air, reinflates the alveoli that collapse under stress. With those sacs open, the long exhale offloads carbon dioxide far more efficiently, and CO2 buildup is one reason stress feels jittery. Huberman credits Jack Feldman's lab at UCLA and Mark Krasnow's lab at Stanford with the mechanistic work on the pattern, first described in the 1930s, and notes that his own lab is running a study with David Spiegel in Stanford psychiatry on deliberate breathing and emotion. He is careful to say those results aren't in yet.
One to three sighs is usually enough. Ten to fifteen cycles, done lying down, can put a person to sleep.
Two cautions come attached. First, heart rate takes 20 to 30 seconds to fall to baseline, and you don't want it faster: the vasovagal faint that hits people who stand up too quickly or see blood is the calming response overshooting. Second, the sigh recruits the parafacial nucleus, the brainstem center that lets you double up breaths while speaking. Because that nucleus sits near the circuits controlling the jaw and tongue, the pattern also tends to relax the face, which may be part of why it helps people hold their tongue when activated.
Huberman's broader point about the sigh is that it sidesteps a hard constraint: it's very difficult to control the mind with the mind when you're highly activated or very tired. Gratitude and intention work in the middle of the arousal range. At the edges, you need a lever in the body.
Acute stress protects you from infection, right up until you finally relax
Here Huberman goes against the popular narrative most directly. Short-term stress, he says, is good for the immune system. Infection is itself a stressor, and the same brain centers that fire the sympathetic chain also activate the spleen, which releases killer cells to hunt bacteria and viruses. Adrenaline, in the short run, is an immune deployment signal.
His evidence is a study published in the Proceedings of the National Academy of Sciences. Participants were injected with endotoxin, a bacterial wall component from E. coli that normally produces fever, nausea, vomiting, and diarrhea. Half had trained in a breathing protocol closely resembling the method popularized by Wim Hof, itself derived from the older Tummo practice: roughly 25 to 30 deep, rapid breaths, an exhale and breath hold of about 15 seconds, repeated for three or four rounds, then a full inhale and hold. That group experienced little to no symptoms.
Huberman is explicit that this kind of cyclic hyperventilation is activating, the reverse of the sigh. Rapid diaphragm movement liberates adrenaline from the adrenals, which is why the breathing feels alerting and why people prone to anxiety may find it triggers a panic response. Cold water does the same thing through a different door. The protocol works because it is stress, deliberately manufactured, and it comes with hard warnings: never near water (shallow-water blackout has killed people), never without a doctor's clearance, and no breath holds for anyone with glaucoma or eye-pressure concerns. Huberman himself uses the breathing rather than ice baths when he feels a tickle in his throat.
The same logic extends to inflammation. A deep cut on a hiking trail triggers swelling that recruits macrophages (or microglia in neural tissue) to clean the site. The response looks bad and is good, in the short term.
And it explains procrastination. Asked constantly about nootropics, Huberman gives an answer that's hard to sell: the most powerful smart drug available is stress. The fear of failure and the impending deadline sharpen cognition, narrow focus, and prime the kind of duration-path-outcome analysis that gets work done. Procrastinators are self-administering the drug at the last possible moment.
The catch is the crash. The familiar pattern of work, work, work, then vacation, then illness is adrenaline dropping and the immune system dropping with it. His rule of thumb for where acute stress ends and chronic stress begins: when you can no longer get what you consider good sleep, you've crossed the line.
Medium-term stress is a capacity problem, and capacity is trainable
Stress that lasts days to weeks, a quarter with too many commitments or a month caring for a sick relative, works differently. Huberman describes his own threshold moment: a website login failed and, in that state, it felt like the most intense thing in the world. He caught it and laughed, but the overreaction told him where he was relative to his limit.
Stress threshold is the ability to cognitively re-regulate what's happening in the body, and raising it is what he means by stress inoculation. The method inverts the usual mind-body advice. Instead of unifying mind and body in a state of calm, you deliberately drive the body into high activation (a sprint, a hard interval on the bike, a cold shower, a round of cyclic breathing) and then practice calming the mind while the body stays flooded with adrenaline. He calls it healthy dissociation, to distinguish it from the clinical disorder.
The lever for calming the mind is visual. Stressed pupils dilate and vision narrows into a tunnel; the eyes are wired into the same cranial-nerve system that governs autonomic state. Deliberately widening the gaze into panoramic vision, without moving the head, releases a brainstem circuit associated with alertness. At 80 or 90 percent of maximum effort, dilating the gaze lets the mind settle while output stays high. Huberman notes the research base here is limited, and that variants of the practice circulate in athletic and military training.
Done once a week or so, he says, what once felt overwhelming starts to feel manageable, with the stressors the same size as before.
Long-term stress yields to serotonin and the absence of tachykinin; oxytocin barely figures
Chronic stress, months and years of elevated adrenaline, is the version that deserves its reputation. Constricted vessels and hypertension drive heart disease, the leading cause of death in the United States. Heart rate variability, the same breathing-heart loop described earlier, is a marker of a system that can still swing between states; a chronically high or chronically low heart rate are both signs of trouble.
The standard prescriptions apply: exercise, sleep, real-time tools. But Huberman says the data point hardest toward social connection, and he wants to strip the new-age vocabulary off that claim and replace it with neurochemistry.
Oxytocin, the hormone credited with every hug and dog pat, is released only in narrow circumstances: post-orgasm, milk letdown, the intense pair bonding of parent and infant. A fist bump doesn't qualify.
Serotonin works on faster timescales and is the better candidate. Seeing someone you recognize and trust releases it, and serotonin supports immune function and synaptic maintenance, the processes that prevent the long-term withering of connections seen in stressed brains. It also produces the feeling of having enough, which is why drugs that raise it can blunt affect and libido as side effects. The behavioral trigger doesn't have to be human. Huberman cites his bulldog, and extends the category to anything that produces delight: play, a pursuit you witness or participate in with real enjoyment. Written gratitude, he adds, appears to act on the same system, though he flags it as the point where his argument shifts from objective to subjective ground.
The other half of the picture is a punishment signal. Tachykinin is a molecule that rises with prolonged social isolation in flies, mice, and humans, and it makes organisms more fearful and paranoid while impairing immune function. Huberman keeps a sticky note reading "tachykinin" above his desk because his own goal-directed tendencies can crowd out the long meals and unhurried company that suppress it. He credits his Stanford colleague Robert Sapolsky for the observation that no primate in history has interacted with so many strangers at a distance while feeling connected to so few.
Connection, he admits, costs something: eating on other people's schedules, staying up later than you'd like, tolerating things not being exactly the way you want them. That flexibility is the price of the neurochemistry.
Supplements that help, and the melatonin trap
For periods when behavioral tools aren't keeping up, Huberman discusses three non-prescription compounds and one to avoid, with the standing recommendation to check every claim on examine.com, a free site that links each supplement's effects to the underlying PubMed studies and specifies which populations were tested. He has no relationship with the site.
Melatonin is the trap. Typical supplemental doses of one to three milligrams are far above physiological levels, and chronic high dosing can suppress cortisol and epinephrine output from the adrenals enough to mimic adrenal insufficiency. It also suppresses puberty in several species, which is his particular concern for children.
That leads to a myth he wants dead: there is no such thing as adrenal burnout. The adrenals hold enough adrenaline for roughly 200 years of stress. The idea traces to Hans Selye, the Nobel laureate who described the general adaptation syndrome and coined eustress and distress; his prediction of a physiological exhaustion phase turned out to be wrong, but the public ran with it. Feeling exhausted after chronic stress is real. The adrenals aren't the reason, unless you have actual adrenal insufficiency syndrome, which is a clinical condition.
L-theanine at 100 to 200 milligrams raises GABA and quiets the ruminating forebrain. He uses it before sleep and, during stretches of poor stress management, in the late afternoon. Examine's summary lists eight studies showing a minor effect on anxiety, four on relaxation, and evidence for reduced task-completion anxiety. That last effect, he notes wryly, may not suit procrastinators who depend on the anxiety.
Ashwagandha has the strongest data for cortisol: across six studies, reductions of 14.5 to 27.9 percent in otherwise healthy but stressed adults, plus five more studies on fatigue and cognitive impairment. It lowers total cholesterol by up to 10 percent, which cuts both ways since cholesterol is the precursor to testosterone, estrogen, and cortisol itself. He takes it only during high-stress periods, never year-round.
Taurine, common in energy drinks now marketed with theanine, is his personal no. He reports it burst microvasculature in his sclera, while acknowledging others swear by it.
Emotion is the gap between your state and the demand
All of this builds toward a definition. Drawing on Lisa Feldman Barrett's work at Northeastern (with the caveat that they don't agree on everything), Huberman proposes that the valence of an emotion, its goodness or badness, comes from whether your internal state matches the demand in front of you.
Being anxious with a lot to do feels bad. Being exhausted with a lot to do feels bad in the same way, because in both cases the state is wrong for the task. Being exhausted at bedtime feels fine. Being wired at bedtime doesn't. The labels we apply (stressed, worried, anxious) matter less than the mismatch underneath them.
On that view, alertness and calm carry no inherent value. You're on a seesaw between them all day, and the only question is whether your position fits the moment. Pop psychology's advice to be responsive rather than reactive is correct and useless without a method; the method is modulating the acute response in real time, before the difficult conversation rather than after a meditation retreat. Huberman even grants permission to close your eyes while receiving hard feedback, since it improves auditory focus.
He is skeptical of mindfulness as an all-purpose tool, at least for himself, because the observer stance pulls him out of the work. What he wants instead is the ability to focus hard until 9:30 and be asleep by 10:02, which requires controlling the stress response with something other than the mind.
The tools he offers are levers attached to real tissue: a diaphragm, a pair of pupils, a spleen, a molecule that punishes solitude. Pull them, and the internal state starts to fit the demand more often. The events outside remain uncontrollable, as they always were.
