Placebo is real biology, and it's dose-dependent

The Short Version

Placebo is not imagination. Expectations formed in the brain's prefrontal cortex travel down real pathways to the hypothalamus, brainstem and body, where they can raise dopamine, shift growth hormone, cortisol, insulin and the hunger hormone ghrelin, lower blood pressure and change how fast a wound heals [1][3][4]. The effect scales with the expectation: people told they got a bigger dose of nicotine performed better and showed more brain activity than people told they got less, even though the actual dose was the same [1][2]. But Andrew Huberman insists the effect has hard limits. A placebo eases the discomfort of cancer treatment but does not shrink tumors, and it eases the discomfort of asthma but does not fix breathing [1]. Ellen Langer goes further than Huberman, calling placebos "probably our very strongest medicine" [3]. Put together, the sources say belief is a real lever that works alongside treatment, not instead of it.

Table of contents

What They Said

Huberman: a prediction machine wired into the body

In a full episode on the subject, Huberman argues against the idea that placebo effects are "simply psychological or the power of the mind over matter." In his words, "it's not that at all placebo, no-sebo, and belief effects actually change the way your biology your physiology works" [1]. He separates three terms. A placebo is an inert substance or behavioral treatment that improves symptoms or performance. A nocebo is an inert treatment that makes them worse. A belief effect happens when specific information about a food, drug, stressor or protocol changes your expectation and therefore the outcome [1]. What they share, he says, is that "they all work by changing expectation" [1].

Expectation lives largely in the prefrontal cortex, which a neurosurgeon guest described to him as the brain structure that controls other brain structures by activating or suppressing them [1]. Some of its subregions have direct connections to the hypothalamus and brainstem, which control heart rate, blood pressure, body temperature, immune function, stress and reward [1]. That, Huberman argues, is how knowing what a pill is supposed to do can change basic physiology. He adds that several highly esteemed researchers believe placebo, nocebo and belief effects "should actually be leveraged in the treatment of various diseases as their own unique treatment" [1].

Dopamine in Parkinson's disease. The example Huberman finds most striking comes from Fabrizio Benedetti's book on placebo effects. Parkinson's involves the loss of dopamine neurons in the substantia nigra, and most effective treatments, such as L-dopa, raise brain dopamine [1]. In placebo-controlled trials, some patients given a sugar pill not only improved but showed genuine increases in dopamine release. Researchers used PET imaging with raclopride, a labeled molecule that binds dopamine receptors. After placebo, less raclopride could bind, meaning more of the brain's own dopamine was occupying the receptors [1].

So why not just give everyone placebo? Huberman gives two reasons. Placebo-driven dopamine increases are generally smaller than those from real drugs. And the strength of expectation is hard to pin down across people: "if people realize they're taking a placebo, the magnitude of the dopamine increase is actually decreased" [1]. That, he says, is exactly why clinical trials use placebo controls, to separate what a drug does from what people expect it to do [1].

Hormones on cue. In another study, people were first taught about growth hormone and cortisol, which didn't change their levels. On days two and three they were injected with sumatriptan, a drug that raises growth hormone and lowers cortisol. Later they were injected with plain saline, and their growth hormone still rose and their cortisol still fell. This happened regardless of what they were told, even when told the injection would lower growth hormone [1]. The common thread was the injection itself, which the body had paired with those hormone changes. Huberman marvels at this, because nobody can simply decide to release growth hormone from the pituitary [1]. He links it to everyday conditioning. Bakery smells can trigger insulin release, and in experiments a buzzer sounded during meals can later trigger an insulin response with no food present. "So are we all just like salivating dogs? I guess we are," he says [1].

Packaging, color and invasiveness. Huberman lists findings that map onto everyday life. A placebo with a brand name, even a made-up one with a trademark symbol, beats a generic one. A pill pushed out of foil or taken from a labeled bottle beats one handed over on a tray [1]. Color matters. When told a pill would help sleep, people given blue placebos slept best. When told it was a stimulant, red worked best. When told it was an antidepressant, yellow brought the most relief [1]. The more invasive or elaborate the intervention, the bigger the effect: capsules beat tablets, injections beat pills, and lying in an inert medical machine that hums and beeps produces the largest effect of all [1]. None of these objects acts on any specific biological system. "In other words, the belief effects, the expectations are real. They are having effects through true biological circuitry. It's just that the pills and the treatments and the machines are not doing anything specific at all except activating expectation" [1].

The dose-dependent nicotine study. Huberman calls this one of his all-time favorite placebo papers. Its title is "Nicotine-related beliefs induce dose-dependent responses in the human brain" [1]. People vaped nicotine, a cognitive enhancer that acts on acetylcholine receptors involved in attention, and were told they were getting a low, medium or high dose. According to the episode summary, everyone actually got the same amount, and researchers checked how much reached the bloodstream [1]. Those told medium performed better on a cognitive task than those told low, and those told high did best. Brain imaging showed activity in the relevant nicotine-responsive region scaling with the dose people believed they'd taken [1]. The lesson: the size of the effect tracks the size of the expectation, as long as the outcome is something expectation can influence, and it shows up in measurable brain activity [1]. He is careful not to endorse vaping. It is probably healthier than smoking but still not good for you, and nicotine raises blood pressure and constricts blood vessels [1].

Mindsets: the milkshake and the housekeepers. From his Stanford colleague Alia Crum's lab, Huberman describes a study titled Mind Over Milkshakes: Mindsets, Not Just Nutrients, Determine Ghrelin Response. People drank the same 380-calorie shake twice, once labeled a 620-calorie "indulgent" shake and once a 140-calorie "sensible" one [1]. When they believed it was indulgent, ghrelin, a hunger hormone secreted by the stomach, dropped more steeply and they felt fuller [1]. In another study, prompted when Crum's advisor suggested that perhaps all the benefits of exercise are placebo, hotel housekeepers were split into two groups. One was told their work was important and valuable. The other was told their daily activities counted as exercise that could lower blood pressure and body weight. Only the second group improved in blood pressure, resting heart rate and body weight [1].

Where the limits come from. Huberman is emphatic: "I do not want to give you the impression that the placebo effect is limitless because it is not limitless" [1]. In cancer care, people told that an inert substance will ease nausea or pain during chemotherapy, radiation or immunotherapy often do feel less pain and nausea and better well-being. "Those are real effects," he says, but "those placebo effects are not acting directly on tumors to reduce their size or eliminate them" [1]. The reason is anatomical: "the outputs of the prefrontal cortex are limited," and "those connections do not extend to tumors themselves" [1]. He separates three things carefully. First, behavioral practices such as stress reduction, better sleep and social support are not placebo but real effects, and the data "have conclusively shown" they can improve cancer treatment outcomes [1]. Second, genuine treatments shrink tumors. Third, placebo effects are real but act on symptoms like discomfort [1].

A study from Ted Kaptchuk's lab at Harvard, published in the New England Journal of Medicine, makes the same point with asthma. When patients were briefly taken off their medication, those given the real drug improved both their breathing and their discomfort. Those given a placebo felt less discomfort, but their actual breathing didn't change [1]. Beliefs are powerful, Huberman concludes, but not omnipotent [1]. The same logic explains why belief can lower blood pressure but can't grow your left quadriceps. As far as anyone knows, there is no direct pathway from prefrontal thoughts to muscle growth, but there are pathways to the hypothalamus and brainstem [1].

The circuit, and why people differ. Huberman describes a rat study, A central master driver of psychosocial stress responses in the rat, which traced a prefrontal subregion to neurons in the dorsomedial hypothalamus. Those neurons produce hallmark stress responses such as higher blood pressure, vasoconstriction and heat production, then signal through the brainstem to the spinal cord, blood vessels and organs. Analogous circuits exist in humans [1]. With a substrate like that, he argues, we shouldn't be surprised by placebo effects at all [1]. People also vary. The first formal placebo study found about 30% of participants had a robust response and 70% a weaker one, and that pattern has recurred [1]. Variation in the COMT gene, which encodes an enzyme that breaks down dopamine, epinephrine and norepinephrine, correlates fairly strongly with how much someone responds to certain placebo conditions [1]. His summary: "It's our thoughts, our mind creating real biological effects" [1].

Huberman in Sydney: "it is dose dependent"

At a live Q&A in Sydney, asked whether he believes in the placebo effect, Huberman answered, "Absolutely" [2]. "So the placebo effect is real. Our belief about what we've taken or what is happening to us has a powerful effect on our physiology. It's not purely psychological" [2]. He adds that even the word psychosomatic is starting to fall away as scientists understand how strongly beliefs affect the body [2].

His first example is Alia Crum's stress research. People who watch a short video about the damage stress does to cognition, sleep and well-being experience those harms. People who watch a video on stress as performance-enhancing get sharper [2]. He calls these belief effects, and draws a distinction: "placebo effects tend to be more general belief effects tend to be around specific types of information" [2].

Then comes the dose point. "So placebo effect has recently been shown to extend to a dose dependent placebo effect," he says, citing the nicotine paper he'd discussed with Peter Attia [2]. His retelling differs a little from the full episode. Here he says people received either zero or a moderate dose, and "People who were told they had a higher dose of nicotine performed better in this cognitive task, when in fact they consumed zero" [2]. In other words, "everyone gets the same dose, either zero or moderate, but depending on what you're told, your performance changes accordingly" [2]. The brain recordings mattered most to him: "the levels of activity in particular areas of the brain that are relevant for cognition, changed according to what the people believe" [2].

His verdict mixes enthusiasm with caution: "I believe in the placebo effect and it is dose dependent and that raises all sorts of scary concerns about the placebo effect, but it's also pretty darn cool" [2]. The practical upshot: "our belief system, including our understanding of the mechanisms that are likely driving certain effects of drugs or protocols or what have you, is going to play a powerful role in whether or not we get the effect that we want" [2]. Along the way he tells of a Nobel laureate who chewed at least six pieces of Nicorette during one meeting to stave off Parkinson's and Alzheimer's, a story he explicitly labels "anecdata," while warning against smoking, dipping or snuffing nicotine [2].

Huberman on food: belief reaches the gut

In an episode on food and mood, Huberman returns to the milkshake study to show that belief can reach organs far from the brain. One group was told their shake was low-calorie and healthy, the other that it was "very decadent high calorie," and the "high calorie" group showed a much larger drop in ghrelin, "the exact same shake given to both groups" [4]. He calls this "top down mechanisms or modulation of our physiology," and compares it to the way obsessive infatuation and love change pain thresholds [4]. "This is yet another example where beliefs or subjective feelings can impact physiology at the level of the periphery because grillin is released in the periphery in the body. And so this is not just the placebo effect" [4].

He adds a caveat about using this on purpose. According to the summary, he says "these belief effects are not about lying to yourself": they require being naive to the information, so you can't simply tell yourself whatever you'd like to believe [4]. The brain-body relationship runs both ways. What you put in your gut shapes how you feel, and what you believe shapes how your body responds [4].

Ellen Langer: mind and body as one thing

Harvard psychologist Ellen Langer has studied what she calls mind-body unity for about five decades. Her position is that mind and body are not two things connected by some mechanism. They are one thing, so wherever you put your mind you also put your body [3]. She objects even to the phrase "mind-body connection," because a connection implies two separate things with something in between [3]. Huberman agrees strongly: "the division of mind and body is one of the greatest mistakes of thinking in psychology and Western medicine that ever existed" [3].

The origin: choices and longevity. Her health research began with a 1970s nursing home study. Residents encouraged to make small choices, such as where to receive visitors, which night to see a movie and caring for a plant, had lived longer 18 months later than residents who had the same things decided for them [3]. Her explanation became the core of her work: your mind is active and your body complies [3].

The chambermaids. Langer tells the housekeeper study as her own. It grew out of her habit of questioning any proclamation, in this case her student Ali's claims about exercise [3]. (Huberman credits the study to Alia Crum and says it was prompted by Crum's advisor [1]. The two accounts seem to describe the same student-advisor pair from opposite ends.) Hotel chambermaids get a great deal of physical activity, but they said they didn't exercise much, and they weren't healthier than others [3]. Langer randomly split them and taught one group that their work is exercise: "Making a bed is like working in this machine at the gym" [3]. "They're not eating any differently, one group from the other. They're not working any harder," she stresses, yet the group that changed its mindset lost weight, showed changes in waist-to-hip ratio and body mass index, and their blood pressure came down [3]. She frames it as a test of nocebo: exercising without realizing it seemed to strip away some of the exercise's effect [3].

Classic placebo findings. Langer cites people told that ipecac, which normally induces vomiting, would stop their vomiting, and it did [3]. She also describes people rubbed with a leaf that either is or isn't poison ivy: "And your body reacts to your thoughts. If you think it's not poison ivy, you don't get the rash. If you think it is poison ivy, you do" [3].

Sleep, vision and wounds. In a sleep lab study, people woke to a clock showing they'd slept two hours more, two hours less or the true amount. "Biological and cognitive functioning follow perceived amount of sleep," she says [3]. In a vision study, she reversed the eye chart so letters grew larger rather than smaller, changing the expectation from "soon I won't be able to see" to "soon I will," and people could see what they couldn't before [3]. Starting the chart a third of the way down had the same effect [3]. In a wound study with graduate student Peter Aungle, participants with a minor wound sat in front of a clock running twice as fast, half as fast or at real speed, and the wounds healed according to clock time rather than actual time [3]. Her lesson for doctors: telling a patient healing will take a month can become a self-fulfilling prophecy. It is better to say we don't really know, some people heal faster, and here is the fastest case I know of [3].

The counterclockwise retreat. Men around 80 lived for about five days in a retreat set up as it had been 20 years earlier, speaking about the past as if it were the present, while a control group simply reminisced [3]. Put in charge of their own luggage and their own lives after being coddled, even the control group improved, and the experimental group improved more: vision, hearing, memory and strength got better, and the men looked noticeably younger [3]. "I'm not interested in describing what is. I'm interested in seeing what may be," she says [3].

Placebo as medicine. Langer's position is the strongest in these sources: "I think that placebos are probably our very strongest medicine" [3]. People get upset when they learn they were given a placebo, but "they should be excited because if the placebo didn't cure you, who cured you, did it yourself" [3]. She believes "placebo has gotten a bad rap," primarily because of pharmaceutical companies, whose drugs have to outperform placebo to reach market [3].

Diagnoses as probabilities. Langer adds a frame the others lack. Every medical finding is a probability for a group, not a prediction for an individual, so a diagnosis and its expected course are not destiny [3]. About chronic illness: "all it means is the medical world doesn't have a fix. It doesn't mean there's nothing can be done" [3]. Her lab uses attention to symptom variability. Researchers phone patients and ask whether a symptom is better or worse than last time, and why. She says this has shown positive results in multiple sclerosis, chronic pain, Parkinson's and stroke, without side effects or stopping medical treatment [3]. Huberman connects her work to neuroplasticity research at UCSF showing that brain maps changed only for the sense subjects were paying attention to, concluding that "awareness is the gate to brain change" [3].

Where They Agree, and Where They Don't

They agree that placebo is biology. Huberman's "the expectations are real" [1], his Sydney line that it's "not purely psychological" [2], his point that belief changes a hormone released in the gut [4] and Langer's "your body reacts to your thoughts" [3] all say the same thing. Measured outcomes back this up in every source: dopamine release, growth hormone, cortisol, insulin, ghrelin, blood pressure, body weight, rashes and wound healing [1][3][4].

They agree that framing and information drive the effect. Brand names, pill colors and injections [1], a video about stress [2], a label on a milkshake [1][4], being told your job is exercise [1][3], a clock on the wall [3] and an eye chart turned upside down [3] all change the outcome by changing what people expect.

They agree the effect scales. The nicotine study is the clearest case of dose dependence: a bigger believed dose brought bigger effects on performance and brain activity [1][2]. The invasiveness ladder (tablet, capsule, pill, injection, beeping machine) [1] and the indulgent-versus-sensible shake [1][4] point the same way.

They disagree on how far it goes. This is the main divide. Huberman draws firm lines: placebo eases the discomfort of cancer treatment but can't shrink tumors, eases asthma discomfort but doesn't change breathing, and can lower blood pressure but can't build muscle, because the prefrontal cortex has a limited set of outputs [1]. Langer is more expansive. She calls placebos "probably our very strongest medicine" [3], suggests believing cancer is a death sentence can make it one, and thinks spontaneous remissions are more common than medicine assumes [3]. She tells of her mother's breast cancer, which had spread to the pancreas, going away [3]. Huberman's framework treats any benefit to cancer outcomes from stress reduction, sleep and social support as a real behavioral effect, not a placebo effect [1]. Langer's view that mind and body are one thing makes her less interested in drawing that line at all [3].

They explain it differently. Huberman explains placebo in terms of circuits: the prefrontal cortex, a path to the hypothalamus and brainstem, catecholamines, COMT [1]. Langer rejects the idea of a mediator between mind and body and describes a single system, "Everything is everywhere," noting that we just lack the technology to notice most of it [3]. Huberman's framework explains the limits. Langer's framework explains why she keeps testing for exceptions.

The details of the nicotine study differ between tellings. In the full episode, Huberman says everyone got the same nicotine dose and was told low, medium or high [1]. In Sydney, he says people got zero or a moderate dose, and some who got zero but were told a higher dose performed better [2]. Both versions support the same conclusion, but readers should know the two retellings don't match.

They agree that honesty matters, though they frame it differently. Huberman notes that knowing you're on a placebo shrinks the dopamine effect [1] and that belief effects "are not about lying to yourself" [4]. Langer's chambermaids weren't deceived: they were given accurate information that their work counted as exercise [3]. Where the sources converge, the belief that works best is one that is true or at least well founded.

Why It Works

A prediction engine with direct lines to the body. The prefrontal cortex evaluates context and predicts outcomes, and some of its subregions have dedicated highways to the hypothalamus and brainstem, which control heart rate, blood pressure, temperature, immunity, stress and reward [1]. A rat study traced one such route from a prefrontal region to the dorsomedial hypothalamus, and from there through the brainstem to blood vessels and organs, producing stress responses [1]. Humans have the most elaborate prefrontal cortex of any animal studied [1].

Real neurochemistry. Placebo raises dopamine release in Parkinson's patients, measured by reduced raclopride binding [1]. The pills and machines do nothing specific. Expectation works "through true biological circuitry" [1].

Conditioning. The body learns to pair cues with physiological responses: an injection with growth hormone and cortisol changes, a smell or buzzer with insulin release [1]. Placebo partly works like Pavlov's bell.

Scaling with expectation. The stronger the believed dose, the stronger the effect on performance and brain activity [1][2]. That's why more elaborate interventions produce bigger placebo effects [1].

Top-down control of peripheral hormones. Belief about calories changed ghrelin released from the stomach [4], which Huberman places alongside love's effect on pain thresholds as top-down modulation [4].

Attention. Langer's work suggests that noticing and mindset change outcomes: perceived sleep drives cognitive and biological functioning [3], and attending to how symptoms vary helps people with chronic illness [3]. The UCSF finding that brain maps change only for what we attend to supports "awareness is the gate to brain change" [3].

Genes and individual differences. Variation in COMT, which regulates the catecholamines involved in many placebo effects, helps explain why some people respond strongly and others weakly [1].

Why the limits exist. Belief can only act through the outputs the brain actually has. There is no known pathway from prefrontal thoughts to tumors or to muscle growth [1]. Placebo acts on symptoms and on systems wired to the hypothalamus and brainstem. It does not act on everything.

Where It Applies, and Where It Doesn't

Where it applies: Symptoms and discomfort, such as pain, nausea during cancer treatment and the discomfort of asthma [1]. Neurochemistry and hormones: dopamine in Parkinson's, growth hormone, cortisol, insulin and ghrelin [1][4]. Cardiovascular and metabolic measures: blood pressure, resting heart rate, body weight [1][3]. Cognition and perceived sleep [2][3], and skin reactions and wound healing in Langer's studies [3]. Huberman notes that "essentially every system within the brain and body has been shown to be susceptible to placebo effects" [1], meaning each can be influenced, not that any one can be cured.

Where it doesn't: Placebo does not shrink or eliminate tumors [1], does not change asthmatic breathing [1] and, as far as anyone knows, cannot build muscle [1]. The dopamine increase from placebo is generally smaller than from real drugs [1]. The effect varies widely: in the first formal study, only about 30% of people had a robust response [1].

Where it weakens: Knowing you're on a placebo shrinks the dopamine effect [1]. Belief effects require genuine expectation, so telling yourself what you'd like to believe doesn't work [4].

Where it can hurt: Nocebo is the same machinery pointed the other way. Crum's stress video made stress harmful for people taught to fear it [2]. Chambermaids who didn't see their work as exercise didn't get its full benefit [3]. A doctor's confident prediction of slow healing may become self-fulfilling [3]. Huberman notes that dose dependence "raises all sorts of scary concerns" [2].

Where it is not a substitute: Belief works alongside real treatment. Langer's symptom-variability approach works without stopping medical treatment [3], and Huberman separates placebo from the proven behavioral benefits of stress reduction, sleep and social support [1]. For any serious illness, keep working with your doctor and use expectation as an addition to care, never a replacement for it.

What To Do With It

1. Learn how your treatments work. Huberman says your understanding of the mechanisms behind a drug or protocol "is going to play a powerful role in whether or not we get the effect that we want" [2]. Before starting a medication, supplement or training plan, learn what it does and why.

2. See your daily activity as exercise. The housekeepers who were told their work counted as exercise improved blood pressure, weight and waist-to-hip ratio without changing what they did [1][3]. Recognize the stairs, carrying and cleaning in your day for what they are.

3. Think of stress as something that can sharpen you. In Crum's work, people shown how stress can enhance performance got sharper, while those taught its harms experienced the harms [2]. When stress hits, remind yourself that it can raise focus and access to memory.

4. Pay attention to how you frame food. The same shake cut ghrelin more when people thought it was indulgent [1][4]. Treat satisfying meals as satisfying rather than describing them to yourself as thin or unfulfilling, and remember that this works through genuine belief, not self-deception [4].

5. Don't let a sleep score drive your day. "Biological and cognitive functioning follow perceived amount of sleep" [3]. After a rough night, don't let a tracker's verdict become a prediction for the day.

6. Track how your symptoms change, not just whether they exist. Langer's approach of noticing whether a symptom is better or worse than last time, and why, has helped people with MS, chronic pain, Parkinson's and stroke [3]. Keep a simple log and look for patterns.

7. Treat prognoses as probabilities. Medical findings describe groups, not you [3]. When a timeline is given, ask what the fastest recovery looked like, as Langer suggests doctors should volunteer [3].

8. Keep expectations honest. Knowing you're taking a placebo reduces its effect [1], and belief effects aren't about lying to yourself [4]. Build expectations from real information about what an intervention can do.

9. Respect the limits. Use belief to ease discomfort and support recovery, but don't expect it to shrink a tumor or open an airway [1]. Pair a positive mindset with the behavioral basics that affect outcomes, such as sleep, stress reduction and social support [1], and with your prescribed treatment.

Sources

1. How Placebo Effects Work to Change Our Biology & Psychology · Huberman Lab
2. LIVE EVENT Q&A: Dr. Andrew Huberman at the ICC Sydney Theatre · Huberman Lab
3. Using Your Mind to Control Your Physical Health & Longevity · Huberman Lab
4. Aummary Of How Foods Nutrients Control Our Moods · Huberman Lab