Andrew Huberman explains the neuroscience of aggression and how to control it
A solo episode of Huberman Lab Essentials in which Andrew Huberman covers the biology, neural circuits, and practical tools related to aggression.
Summary
Andrew Huberman, neuroscientist and professor at Stanford School of Medicine, presents a comprehensive overview of the science of aggression — covering its neural circuits, hormonal drivers, genetic factors, and practical tools for modulation. He distinguishes between reactive, proactive, and indirect aggression, arguing that these are biologically distinct. A central and counterintuitive claim is that it is not testosterone but estrogen — specifically testosterone aromatized into estrogen in the brain — that activates the neural circuits responsible for aggressive behavior. He further argues that cortisol levels and serotonin levels are the key contextual variables that determine whether those circuits fire, and that day length, sunlight exposure, and supplementation can meaningfully shift an individual's predisposition toward or away from aggression.
Key Takeaways
FULL TRANSCRIPT
Introduction: Types of Aggression and Why They Matter
Andrew Huberman: Welcome to Huberman Lab Essentials, where we revisit past episodes for the most potent and actionable science-based tools for mental health, physical health, and performance. I'm Andrew Huberman, and I'm a professor of neurobiology and ophthalmology at Stanford School of Medicine.
Today we are discussing aggression. I'm going to explain that there are several different types of aggression. For instance, reactive aggression versus proactive aggression — meaning sometimes people will be aggressive because they feel threatened, or they are protecting those that they love who also feel threatened. There's also proactive aggression, where people go out of their way to deliberately try and harm others. And there is indirect aggression, which is aggression not involving physical violence — for instance, shaming people and things of that sort.
It turns out that there are different biological mechanisms underlying each of the different types of aggression. Today I will define those for you. I'll talk about the neural circuits in the brain and body that mediate each of the different kinds of aggression, and talk about some of the hormones, peptides, and neurotransmitters involved. I promise to make it all accessible to you. Even if you do not have any biology or science background, I'm certain that by the end of the episode you will come away with a much more thorough understanding of what this thing that we call aggression really is. When you see it in other people, I think it will make more sense to you. And when you observe it in yourself — the impulse to engage in aggression, verbal or physical or otherwise — I hope that you'll understand it better as well. The tools I will describe should allow you to modulate and control aggressive tendencies or predispositions to aggressiveness, and generally be able to engage with people in a more adaptive way.
Overall, the context of aggression really matters. There are instances where aggression is adaptive — for instance, a mother protecting her children. Of course, other forms of aggression, like unprovoked proactive aggression where somebody is simply being violent to somebody else even when unprovoked, most of us cringe when we see that kind of behavior. It can even evoke aggression in people when they observe it.
Debunking the "Aggression Is Just Sadness" Myth
Many of you have probably heard the statement — which I believe arises from pop psychology, not from formal academic psychology — that aggression is just sadness, a form of sadness that's amplified and shows up as aggression. But when we look at the underlying biology and the peer-reviewed literature on this, nothing could be further from the truth. We have distinct circuits in the brain for aggression versus grief and mourning. Those are non-overlapping. Now, that doesn't mean that you can't be sad and aggressive, or in a state of mourning and aggressive, at the same time. But the idea that sadness and aggression are one and the same thing is simply not true.
By understanding that — or perhaps by understanding that irritability and aggression are not the same thing — you'll be in a much better position to apply some of the tools we will talk about in this episode in order to reduce, eliminate, or, if it's adaptive to you, modulate aggression. And yes, there are cases where modulating your aggression — in some cases even amplifying aggression — can be adaptive.
Konrad Lorenz, Fixed Action Patterns, and the Hydraulic Pressure Model
One of the names most associated with the formal study of aggression is none other than Konrad Lorenz. Konrad Lorenz studied so-called imprinting behaviors and fixed action pattern behaviors — patterns of behavior that could be evoked by a single stimulus. The idea that you can get a whole category of behaviors, like looking to somebody for comfort and only them, triggered by just the presence of that person, is remarkable. Because what it suggested — and what turns out to be true — is that there are neural circuits, not just individual brain areas, but collections of brain areas that work together to engage a pattern of behaviors.
That's the first fundamental principle we need to define today: when we talk about aggression, we're talking about activation of neural circuits, not individual brain areas, but neural circuits that get played out in sequence like keys on a piano. That playing out in sequence means that aggression is a verb. It has a beginning, a middle, and an end. It's a process, not an event. And as you'll see, that turns out to be very important in terms of thinking about how one can halt aggression, prevent it from happening before it's initiated, or maybe even prolong aggression if that's what's needed.
Now, Konrad Lorenz had no real knowledge of neural circuits. He knew there was this thing we call a brain and a nervous system, and he knew that there were chemicals in the brain and hormones and things of that sort that were likely to play a role, but he didn't take any measures to define what the neural circuits were. But he did think about what sorts of underlying processes could drive something like aggression. He talked about one particular feature that's especially important — this notion of a pressure. The idea that certain hormones will bias somebody or an animal to be aggressive, that certain neurotransmitter states will bias somebody to be more or less aggressive, and that historical features based on childhood, etc., will contribute. He understood that there would be a constellation of things that would drive people to be aggressive.
He described a so-called pressure — almost like a hydraulic pressure. Think about fluid pressure in a small container being pushed and pushed until the container is ready to explode, and how multiple variables could impinge on that and create that pressure. It turns out that's exactly the way the system works. There is no single brain area that flips the switch for aggression, although we'll soon talk about a brain structure that generally houses the propensity and the output of aggression.
This notion of a hydraulic pressure that can drive us toward aggressive behavior — or conversely can be very low and keep us in a state of non-reactivity, maybe even passivity or submissiveness — is a very important feature because it really captures the essence of how neural circuits work when we're talking about primitive behaviors generally. You can start to notice this in yourself and in others. You can start to notice when you are veering toward aggression or when someone is veering toward aggression, verbal or physical. That veering is the buildup of this hydraulic pressure that Lorenz was referring to, and it really does have an underlying biological basis.
Walter Hess and the Discovery of the Ventromedial Hypothalamus
It was some years later that the first experiments came along which really started to identify the brain areas and the biological pressures that can induce aggressive behavior. The person who really gets credit for this is a man by the name of Walter Hess, who at that time was working on cats. What Hess did was work with cats that were awake, and he was able to lower a stimulating electrode into their brain. Keep in mind that the brain does not have any pain sensors, so after a small hole is made in the skull, electrodes are lowered into the brain — this is what's done commonly in human neurosurgery. He was trying to identify brain regions that could generate entire categories of behavior.
Eventually, his electrode landed in a site and he provided electrical stimulation that caused this otherwise passive, purring, relaxed cat to suddenly go into an absolute rage. When he turned off the stimulation of this particular brain area, the cat very quickly — within seconds — went back to being a passive, calm kitty. Later experiments done in mice but also in humans confirmed that stimulation of this brain area evoked not just behavioral aggression but also subjective feelings of aggression and anger.
So what was this incredible brain area? The so-called VMH, or ventromedial hypothalamus. The ventromedial hypothalamus is a nucleus — meaning a small collection of neurons — only about 1,500 neurons on one side of your brain and a matching 1,500 neurons on the other side. That combined 3,000 neurons or so is sufficient to generate aggressive behavior of the sort that Hess observed in the cat. And believe it or not, when you see somebody who is in an act of rage, or in an act of verbal aggression, or in an act of defensive aggression protecting their family or loved ones or country, almost certainly those neurons are engaged in that behavior.
David Anderson's Lab and the Estrogen Receptor Neurons
Experiments done by David Anderson's lab at Caltech were really the first to parse the fine circuitry and to show that the ventromedial hypothalamus is both necessary and sufficient for aggressive behavior. What they did was identify where the ventromedial hypothalamus was in the mouse — that was pretty straightforward, as it was known before they started these experiments. Then they analyzed which genes — meaning which DNA, which of course becomes RNA, and RNA becomes protein — which proteins are expressed in particular cells of the ventromedial hypothalamus. It turns out that there's a particular category of neurons in the ventromedial hypothalamus that make an estrogen receptor, and it is those neurons in particular that are responsible for generating aggressive behavior.
How did they know this? They used a tool that's been described by a previous guest of this podcast. We had an episode with the psychiatrist and bioengineer and my colleague at Stanford School of Medicine, Karl Deisseroth. He and others have developed tools that allow people to control the activity of neurons essentially by remote control, by shining light on those neurons. In the context of an experiment on a mouse — and these were the beautiful experiments of Dayu Lin, who is now in her own laboratory at New York University — a little fiber optic cable was put down into the hypothalamus of the mouse. The mouse is able to move around in its cage freely, even though it has a little tether — a very thin wire, what we call an optrode. The experimentalist, in this case Dayu Lin, was able to stimulate the turning on of a little bit of blue light. That blue light activated only those estrogen receptor neurons in only the ventromedial hypothalamus. She was able to do that because she had introduced a gene developed by Karl Deisseroth that allows light to trigger electrical activity in those neurons.
If any of that is confusing, here's the experiment in plain terms. There's a mouse in a cage with a little wire coming out of its head. It doesn't notice — we know this because it's still eating and mating and doing all the things that mice do on a daily basis. The mere pressing of a button activates a little bit of light released at the end of that wire. That light activates particular neurons — in this case, the estrogen receptor-containing neurons in only the ventromedial hypothalamus.
A large number of experiments were done, but the first was to put the male mouse in with a female mouse who is in the so-called receptive phase of estrus — that is, she will allow mating. He starts mating with her, and they go through the standard repertoire of mating behaviors observed in mice. Then, about halfway through the behavior, Dayu Lin turned on the light to stimulate these estrogen receptor-containing neurons only in the male mouse. What she observed was incredibly dramatic. The male mouse ceases from trying to mate with the female mouse and immediately tries to kill her. He starts attacking her. Then she turns off the light. The male stops and goes back to trying to mate with the female mouse. These are such dramatic shifts in behavior triggered only by the activation of only this small set of neurons within the ventromedial hypothalamus. The shift in behavior is almost instantaneous — occurring within seconds if not milliseconds.
The next experiment she did was to put a male mouse with this stimulation capability into a cage alone, but with a rubber glove filled with air or water. She stimulates the activation of these ventromedial hypothalamus neurons and the mouse immediately tries to kill the glove. It goes into a rage attacking the glove as if it were another mouse or some other animate object — but of course it's just a rubber glove. She stops the stimulation and the mouse immediately goes back to being completely calm.
The Periaqueductal Gray and the Physical Outputs of Aggression
Subsequent experiments done by Dayu Lin in her own laboratory and other laboratories have shown that the ventromedial hypothalamus is connected with a bunch of other brain areas. One I want to call out now is the so-called PAG, the periaqueductal gray nucleus. This is a large structure in the back of the brain that houses neurons that can create opioids. These are neurons that can produce endogenous — meaning made by the body — chemicals that can cause pain relief. You could understand why that might occur in a circuit for aggression: even if one is the aggressor, it's likely that they may incur some physical damage and they'd want some pain relief. The PAG is also connected to a number of neural circuits that eventually arrive at things like the jaws. In fact, stimulation of the ventromedial hypothalamus can evoke aggressive biting behavior.
Aggressive biting behavior is particularly interesting because in humans, and especially in human children, biting is something that while young children might do as a form of aggression, tends to disappear pretty early in childhood. If it doesn't, it's often seen as a mark of pathology. There is general agreement in the psychology and psychiatric community that past a certain age, using one's teeth to impart aggression and damage on others is a particularly primitive and troubling behavior.
Dayu Lin's lab has shown that activation of the ventromedial hypothalamus triggers a downstream circuit in the periaqueductal gray, which then triggers a whole other set of circuits of fixed action patterns — here we are back to Lorenz — including swinging of the limbs, punching, and biting behavior. It's remarkable that we have circuits in our brain that can evoke violent use of things like our mouth or violent use of things like our limbs, which of course could be used for things like singing or kissing or eating or gesticulating in any kind of polite or impolite way.
The point here is that neural circuits — not individual brain areas — evoke the constellation of behaviors that we call aggression.
Testosterone, Estrogen, and the Aromatization Surprise
Now, many of you are probably puzzled, because I've been talking about this highly specialized brain area, the ventromedial hypothalamus, and this highly specialized subcategory of neurons — the neurons that make estrogen receptor — and yet the activation of those cells triggers dramatic and immediate aggression both in males and in females, and both against males and against females. So what's going on here?
Most of us think about estrogen and we don't immediately think of aggression. Most of us hear testosterone and we might think about aggression. To make a long story short and to dispel a still unfortunately very common myth: testosterone does not increase aggressiveness. Testosterone increases proactivity and the willingness to lean into effort in competitive scenarios. If people are given testosterone, or if you look at people who have different levels of testosterone that they naturally make, what you'll find is that testosterone tends to increase competitiveness — but not just in aggressive scenarios. So if somebody is already aggressive, giving them testosterone will have the tendency to make them more aggressive. If somebody, however, is very benevolent and altruistic, giving them testosterone will make them more benevolent and altruistic, at least up to a point.
It turns out there's evidence that in certain contexts estrogen can make people more aggressive. So what's going on? Well, testosterone can be converted into estrogen through a process called aromatization. There's an enzyme called aromatase — anytime you have a word that ends in "-ase" in the context of biology, it's almost always an enzyme. The aromatase enzyme converts testosterone into estrogen, and it is actually testosterone aromatized — converted into estrogen — and then binding to these estrogen-containing neurons in the ventromedial hypothalamus that triggers aggression.
I want to repeat that: it is not testosterone itself that triggers aggression. It is testosterone aromatized into estrogen within the brain, and binding to these estrogen receptor-containing neurons in the ventromedial hypothalamus, that evokes aggression — and dramatic aggression at that.
This effect of estrogen causing aggression in the brain is very robust. So much so that if you take a mouse that lacks the aromatase enzyme — or a human that lacks the aromatase enzyme, and they do exist — then there is a reduction in overall aggression despite high levels of testosterone. It doesn't matter how much you increase testosterone or any of its other derivatives; you do not observe this aggression. This runs counter to everything that we know and think about the role of testosterone. So the next time somebody says testosterone makes people aggressive, you can say: actually, it's estrogen that makes people aggressive — and animals aggressive, for that matter.
Now, of course, because males have relatively less estrogen circulating in their brain and body than females — because they have testes, not ovaries — testosterone is required in the first place in order to be converted into estrogen to activate this aggressive circuit involving these estrogen receptor-containing neurons in the ventromedial hypothalamus. So anytime you hear that testosterone is high, you should think: testosterone is high in the body, and perhaps estrogen is low in the body, but that means there's going to be heightened levels of estrogen in the brain and therefore an increased propensity for aggression.
In females who generally make less testosterone relative to estrogen, there is sufficient estrogen already present to trigger aggression. So both males and females are primed for aggression. But that's riding on a context, and that context — whether or not you get a tendency for aggression — depends on whether or not cortisol is high or low.
Day Length, Cortisol, Serotonin, and the Internal State
There are beautiful data showing that whether or not estrogen stimulates aggression can be powerfully modulated by whether or not days are short or days are long — in other words, whether or not there's a lot of sunshine or not. Day length is converted into hormonal signals and chemical signals. The primary hormonal and chemical signals involve melatonin, dopamine, and also the stress hormones.
To make a very long story short: in long days, where we get a lot of sunlight both in our eyes and on our skin, melatonin levels are reduced. Melatonin is a hormone that tends to produce states of sleepiness and quiescence. It also tends to activate pathways that reduce things like breeding and sexual behavior. In long days, dopamine is increased — dopamine being a molecule associated with feelings of well-being and motivation and the desire to seek out all sorts of things. And in long days, provided we're getting enough sunlight on our skin and to our eyes, the stress hormones — especially cortisol — are reduced in levels.
If estrogen levels are increased experimentally under long day conditions, it does not evoke aggression. However, in short days, if estrogen is increased, there is a heightened predisposition for aggression. That makes perfect sense. Short days tend to be associated with winter. In winter, we are bombarded with more bacteria and viruses because bacteria and viruses actually survive better in cold than in heat. So shorter days are conducive to aggression — not because days are short per se, but because stress hormone levels are higher and dopamine levels are lower.
Here's where all of this starts to converge on a very clear biological picture, a very clear psychological picture, and indeed a very clear set of tools. Under conditions where cortisol is high — where the stress hormone is elevated — and under conditions where the neuromodulator serotonin is reduced, there is a greater propensity for estrogen to trigger aggression.
For males who make a lot of testosterone relative to estrogen: anytime testosterone is high in the body, there is going to be some aromatization — that conversion of testosterone to estrogen. So anytime testosterone is high, there's going to be heightened levels of estrogen in the brain and therefore an increased propensity for aggression. In females who generally make less testosterone relative to estrogen, there is sufficient estrogen already present to trigger aggression. So both males and females are primed for aggression, but whether that tendency fires depends on whether cortisol is high or low.
I'm telling you that if cortisol is relatively higher in any individual, there's going to be a tilt — an increase in that hydraulic pressure that Lorenz talked about — toward aggression. And if serotonin, the neuromodulator associated with feelings of well-being and sometimes even slight passivity, is low, there's also going to be a further shift toward an aggressive tendency.
So if we return to Lorenz's hydraulic pressure model of aggression, we realize that external stimuli — things that we hear, things that we see, for instance someone saying something upsetting or us seeing somebody do something that we don't like — as well as our internal state, our subjective feelings of well-being, our stress level, our feelings of whether or not we have enough resources and are content with what we have — all of that is converging on this thing we call internal state and creating this pressure of either being more aggressive or less aggressive.
We have some major players feeding into that final pathway — the question of whether or not we will hit the other person, whether we will say the thing that is considered aggressive. There are many things funneling into that question, but we can really boil them down to just a few common elements. Those elements are whether or not cortisol levels are relatively lower or relatively higher. Relatively higher is going to tend to make people more reactive. Why? Because reactivity is really a function of the autonomic nervous system, which oscillates between the so-called sympathetic arm — which tends to put us into a state of readiness through the release of adrenaline — and the parasympathetic arm. Cortisol and adrenaline, when they're circulating in the brain and body, make us more likely to move and to react and to speak. They will actually induce a kind of low-level anticipatory tremor — a body in motion is more easily set into further motion.
Practical Tools for Reducing Cortisol and Aggressive Tendency
In terms of keeping cortisol in a range that's healthy and doesn't bias someone toward high levels of aggression and irritability, that's going to be set by a number of larger modulators or contextual cues. Obviously, getting sunlight in your eyes early in the day and as much sunlight as you safely can throughout the day is going to be important — again, because of this effect of estrogen in long days not increasing aggression, whereas in shorter days estrogen increases aggression because of the increase in cortisol observed in short days.
Another way to reduce cortisol involves the use of sauna and heat, but also hot baths. It turns out that hot baths and sauna can be very beneficial for reducing cortisol. To give a synopsis: a 20-minute sauna at anywhere from 80 to 100 degrees Celsius is going to be beneficial for reducing cortisol. If you don't have access to a sauna, you could do a hot bath.
And of course, some of you may be interested in exploring the supplementation route. For reductions in cortisol, the chief player there is ashwagandha, which is known to decrease cortisol fairly potently. I should warn you that if you're going to use ashwagandha in order to reduce cortisol, first check with your doctor or healthcare provider before adding or subtracting anything from your supplementation or health regimen. Chronic supplementation with ashwagandha can have some not-so-great effects — disruption of other hormone pathways and neurotransmitter pathways. The limit seems to be about two weeks of regular use before you'd want to take a break of about two weeks. So ashwagandha is a very potent inhibitor of cortisol but with some other effects as well — don't use it chronically for longer than two weeks.
But if your goal is to reduce cortisol — let's say you're going through a period of increased irritability and aggressive tendency, maybe you're also not getting as much light as you would like, and perhaps there are other circumstantial things leading you toward more aggressiveness — ashwagandha can be potentially helpful.
Genetic Variants in Estrogen Receptor Sensitivity
In light of all this about cortisol, estrogen, and day length, I should mention that there are in fact some people who have a genetic predisposition to be more irritable and aggressive. There is a genetic variant present in certain people that adjusts their estrogen receptor sensitivity, and that estrogen receptor sensitivity can result in increased levels of aggression — sometimes dramatic increases. However, and also very interestingly, photoperiod — meaning day length — is a strong modulator of whether or not that aggressiveness turns up or not. Whether or not that person with the particular gene variant is more aggressive depends on how long the day is and how long the night is.
One particular study that I like which references this is Trainor et al. The title of the study is "Photoperiod Reverses the Effects of Estrogens on Male Aggression via Genomic and Non-Genomic Pathways." This was a paper published in the Proceedings of the National Academy of Sciences. It really points to the fact that rarely is it the case that just one gene will cause somebody to be hyperaggressive. Almost always there's going to be an interplay between genetics and environment. As environment changes — such as day length changes and the length of night changes — so too will the tendency for people with a given genetic variant to be more aggressive or not.
Now, in the absence of detailed genetic testing for this particular estrogen receptor variant, most people are probably not walking around knowing whether they have this gene or not. Regardless, I think it's important to pay attention to how you feel at different times of year depending on whether it's summer or winter, whether or not you're getting sufficient sunlight, whether or not you're getting sufficient sunlight exposure to your skin, whether or not you're indoors all the time. Generally those things correlate with season, but not always. You can go through long bouts of hard work in the summer months when days are long but you're indoors a lot and getting a lot of fluorescent light exposure late in the evening, and perhaps that's when you're feeling more aggressive. So we have to be careful about drawing a one-to-one relationship between any biological feature and any psychological or behavioral feature like aggressiveness.
But it is helpful to know that these genetic biases exist. How they play out — they shift our biology in a general thematic direction. They don't change one thing; they change a variety of things that bias us toward or away from certain psychological and behavioral outcomes. And we described earlier the various things we can do in order to offset them — in terms of trying to keep cortisol low by getting sufficient sunlight regardless of time of year and regardless of whether or not you happen to have this particular genetic variant.
Acetyl-L-Carnitine, ADHD, and Reducing Aggressive Behavior
I want to share with you a study that's focused on kids but that has important ramifications for adults as well. There are many kids out there that suffer from so-called attention deficit hyperactivity disorder, or ADHD. There are also many adults we are finding who are suffering from ADHD. The study I'm about to share with you explored how a particular pattern of supplementation in kids with ADHD was able to reduce aggressive episodes and impulsivity and increase self-regulation. The title of the study is "Efficacy of Carnitine in the Treatment of Children with Attention Deficit Hyperactivity Disorder." What they focused on was whether or not acetyl-L-carnitine supplementation could somehow adjust the behavioral tendency of these kids with ADHD. And to make a long story short, indeed it did.
There was a very significant effect of acetyl-L-carnitine supplementation on improving some of the symptomology of ADHD. This was a randomized double-blind placebo-controlled double crossover study. They showed significant reductions in their so-called total problem score — a well-established measure of behavioral problems in kids with ADHD, and I should say adults with ADHD as well. Reductions in attentional problems, overall reductions in delinquency, and most important for the sake of today's discussion, significant reductions in aggressive behavior. They were able to confirm the shifts in acetyl-L-carnitine within the bloodstream of these kids — that is, they were able to correlate the physiology with the psychological changes.
Studies such as this are useful because they point to the fact that very seldom, if ever, will there be one supplement or one nutritional change or even one behavioral change that's going to completely shift an individual from being aggressive and impulsive. Rather, by combining different behavioral regimens, by paying attention to things like time of year and work conditions and school conditions and overall levels of stress and likely therefore levels of cortisol, you can use behaviors, diet, and supplementation as a way to shift that overall internal milieu from one of providing a lot of internal hydraulic pressure — as it's been called throughout this episode — toward aggressive impulsivity, and instead relax some of that hydraulic pressure and reduce aggressive tendencies.