The Driver

Even single-celled organisms have threat detection built in. If you put a single bacterium in a dish with poison on one side and food on the other, it will move toward the food and away from the poison. Which makes sense when you realize that aside from reproduction, the ability to avoid threat is the most important survival function any living being has. It is so vital that something very close to our human threat detection system exists in the lamprey, a jawless fish that's been around for 500 million years and is one of our earliest vertebrate ancestors. The pressure to get it right was so intense that evolution perfected it almost immediately, and it has remained largely unchanged since. To put 500 million years into perspective, the first dinosaurs appeared 230 million years ago and went extinct 66 million years ago. This fish predates them by almost 300 million years.

The lamprey's brain is so primitive it barely looks like a brain at all; it looks more like a swollen nerve cord. But when researchers examined this ancient blob of neural tissue, seemingly so different from modern animals, they found the same threat detection architecture that we see in modern humans today. An amygdala-like structure for detecting danger, a periaqueductal gray (PAG) for coordinating response, and a hypothalamus for triggering threat chemistry. This basic setup appears in every vertebrate, and has for half a billion years. And what we see in our world today are the winners of this long evolutionary march toward the present. Anyone whose threat detection system wasn't quite up to snuff? They're not our ancestors, or anyone's ancestors. Their genes got edited out of the evolutionary line.

Half a billion years is a long time not to get an upgrade. Evolution tinkers constantly - unless something is working so well there's nothing to improve. So let's look at how this system actually works, and why evolution never found a reason to change it.

This brain system's whole job is to keep us alive by evading threats, and it does so in a one-two-three punch that happens in the blink of an eye. Which makes evolutionary sense, because the animals who could respond the fastest were the ones that got to live. It really is as simple as that. Part one of this sequence happens in the amygdala, whose job it is to process your surroundings and decide what’s a threat. It has two main ways of doing this: the high road, which involves conscious thought and that we will get into more later, and the low road, which is more relevant for this evolutionary snapshot we’re painting right now. The way the low road works is that sensory information, what you see, hear, smell etc., gets sent directly to the amygdala before your conscious mind even knows it's there. This takes about 12 milliseconds. The amygdala gets a rough, blurry sketch of what's happening - something big moving fast, a loud sudden noise, a shape that pattern-matches to "predator" - and it fires. You jump before you know what scared you. You flinch before you've decided to flinch. The whole point is speed over accuracy, because for 500 million years, the cost of reacting to a false alarm was low, and the cost of reacting too slowly was death.

And because overreacting is evolutionarily better than underreacting, the amygdala is primed to overattribute things that are similar to threats as threats. Remember the woman from the last chapter who had a full asthma attack upon seeing a paper rose? Her amygdala had learned "rose = danger" so thoroughly that the visual pattern alone triggered the entire cascade. No conscious thought required, no time to say "wait, that's paper." By the time the high road could have evaluated the situation, her body was already reacting. This is what’s happening with ever-expanding lists of food sensitivities, or allergies. And this is the system at play in many of the diseases we know about.

Part two of the threat response is the job of the periaqueductal gray, otherwise known as the PAG, which coordinates the body’s actual response. The PAG has two primary modes of response, which you can think of almost like a toggle switch. The brain quickly asks the simple question: will action improve this situation? If the answer is yes, if there's somewhere to run, something to fight, some action that might work, the dorsolateral part of the PAG activates, which we’ll call the mobilizing PAG. Your heart rate spikes, your energy stores are mobilized, and your muscles prepare for action. This is fight or flight, the one everyone's heard of.

But if the answer is no, like if you're trapped, or if the threat is too big, or if there's just nothing you can do, a different part activates. The ventrolateral PAG triggers freeze, shutdown, and/or collapse, which we’ll call the immobilizing PAG. Your heart rate drops, your body starts trying to conserve energy, and in extreme cases, you can dissociate, or go numb. The classic example is prey animals playing dead when caught by a predator - that's the one most people have heard of. But the freeze response isn't just for "a predator has you in its jaws." Sometimes the answer is "not right now, but maybe later." An animal that detects a predator but hasn't been seen yet holds perfectly still. The bet is: stay frozen, and maybe when the threat passes, I can escape. Sometimes the answer is "no, and fighting would make it worse." When a lower-ranking animal encounters a dominant one, it shows submission and withdrawal rather than fighting a battle it would lose. Better to back down than get torn apart. The system is conserving resources for situations where action might actually help.

Sometimes the answer is just "no." When an animal is sick or injured, it hunkers down and waits. Or when an animal is subjected to inescapable threat - threat where nothing it does changes the outcome - it eventually stops trying at all. Researchers documented this with dogs given inescapable shocks. The dogs would struggle at first, trying everything to escape. But after enough exposure to "nothing works," they'd just... stop. Even when escape became possible later, they wouldn't take it. The system had learned: action doesn't help here. Stop wasting energy on it. Even fish do this.

And people have been noticing these two opposite patterns for decades. Attachment theory identified anxious attachment and avoidant attachment - which map perfectly onto people whose PAG tends to flip one way versus the other more often based on what their systems learned early on. Polyvagal theory talks about dorsal vagal shutdown and sympathetic activation which are again, these two responses playing out. What both were witnessing was this very clear PAG switch pattern; these patterns are obvious if you watch human behavior. Or really any animal's behavior. Because this toggle between "I can act" and "I can't act" is the oldest decision in the animal playbook.

Part three is the job of the hypothalamus, which you can think of as the pharmacy of the brain. Within seconds, your hypothalamus releases the entire biochemical cascade that shifts your body from homeostasis to survival mode. Your heart rate changes, your blood flow redirects, digestion shuts down, glucose mobilizes. Everything about your body's chemistry reorganizes around one goal: survive this.

And for 500 million years, this system worked perfectly. A threat would appear, the amygdala would detect it, the PAG would coordinate the response, and the hypothalamus would deploy the chemistry to fuel it. The animal would fight, flee, or freeze, and then the threat would pass. And once this happened, the system would reset. The chemistry flooding the body would start to clear, heart rate would return to baseline, digestion would turn back on, and the animal would go back to living normally. The threat response was designed to last minutes, maybe hours in extreme cases, and then it was over. And this is still how it works, in every animal on the planet. The ancient system activates when needed and deactivates when the threat resolves.

Except in one species.

Because around 100,000 years ago, a different switch got flipped. For hundreds of thousands of years, anatomically modern humans existed with brains very similar to ours. They made the same basic stone tools, generation after generation. They lived in small groups, and left little trace of symbolic thought.

Then suddenly, the archaeological record explodes with evidence of radical behavioral change. Cave paintings appear depicting animals, humans, and abstract symbols. People start wearing jewelry made from shells and bones, some transported hundreds of miles from their origin. Burial sites begin including grave goods like tools, ornaments, and food, suggesting belief in an afterlife or at least symbolic thinking about death. Musical instruments appear, complex multi-part tools, the creation of which required planning several steps ahead, trade networks spanning continents, and boats capable of reaching Australia across open ocean.

This wasn't evolution in the way Darwin understood it, with gradual changes over millions of years. Harari in Sapiens described this as an almost overnight phenomenon around 70,000 years ago. But the skull record seems to tell a slightly different story - brain size had already reached modern levels by 300,000 years ago, but brain shape continued evolving, with the frontal and parietal regions expanding into their modern globular form between about 100,000 and 35,000 years ago. But whether you call it practically overnight like Harari claims, or something that happened over the course of 65,000 years, both are lightning fast for something this advanced to come online.

And it makes sense evolutionarily why this trait got selected for so quickly. Recursive language alone - our ability to nest ideas within ideas indefinitely, to say not just "danger" but "the man who saw the lion that killed the hunter from the neighboring tribe is afraid to go near the watering hole where it happened" - was an immediate survival advantage. The populations with this capability rapidly outcompeted and replaced those without it.

The mismatch

But because this adaptation happened so fast, there wasn't time for evolution to integrate these new cognitive abilities with the ancient survival systems. A hundred thousand years is nothing in evolutionary time. These kinds of huge biological adaptations typically take a million or more years to develop and refine. We developed it in less than 10% of that time. And it’s only been another 30 thousand or so years since. This is a really new evolutionary ability that came online practically overnight, and we haven’t had time to adapt.

We have the ability to imagine infinite futures, to rehash past threatening events, to have an opinion about what our body is signaling to us, and we keep running all of it through threat detection systems that are still shockingly similar to that of the lamprey. And this mismatch is the core problem. We can now maintain threat activation through thought alone, which is something no other animal can do.

Remember we said the amygdala has two pathways - the low road that reacts in 12 milliseconds, and the high road that involves conscious thought. The low road kept us alive for 500 million years. But the high road? The cognitive revolution gave us the ability to generate threats through thought itself. To remember past dangers, imagine future ones, and override the signals our body is sending.

The lamprey doesn't have a high road. It can't think about the predator that attacked it last week. It can't worry about whether there might be a predator around the next rock. It doesn’t feel the activation and ignore it because it doesn’t believe the signal. It detects threat, responds, the threat passes, and its system resets. But humans? The high road can feed thoughts into the low road, and the low road reacts to those thoughts exactly as it would to a real predator. The amygdala fires, the PAG coordinates, the hypothalamus floods the system. We're triggering the ancient hardware with our thoughts, and it responds the same way it has for 500 million years. And we are the only animal that is able to have an opinion about our threat system’s reactions. We interpret the signals rather than just reacting on instinct. We can react to the threat response itself as a threat, we can ignore it and try to make it go away, we can come up with complex narratives that have nothing to do with the signal our body is sending. Our modern thinking brains never got the chance to evolve to interface effectively with the animal threat detection system, and that mismatch gave us the mechanism to get mentally stuck. And, subsequently, sick.

This is the human condition. Our ability to react and respond to our threat activation with our thoughts. The feedback loop of our thoughts alone triggering our ancient threat system. This interference causing and maintaining pathostasis is why we suffer in ways that animals do not. A zebra runs from a lion, escapes, and goes back to grazing. It doesn't lie awake replaying the attack, doesn't develop a generalized fear of open plains, doesn't notice the lion and ignore the signal its body is sending. The threat passes and the system resets. But humans? We can suffer for decades from things that happened once, from things that might never happen, from purely imagined scenarios our thinking brains generate and our ancient hardware can't distinguish from real danger.

We can see this in reverse when animals are exposed to us. Dogs co-evolved specifically to attune to human nervous systems, and researchers have found their long-term cortisol levels synchronize with their owners' cortisol levels. The owner's personality predicts the dog's threat hormones. Cats in the same households, less selected for human emotional attunement, show no such synchronization. And the disease rates bear this out: dogs get cancer at almost double the rate of cats. Zoo animals, trapped in inescapable confinement, whose immobilizing PAG gets stuck on, develop chronic diseases at rates far exceeding wild populations. Almost one in two captive wolves die of cancer while wild wolves almost never do. The threat system works fine when threats resolve. It's the chronicity that breaks it - and we're the first species with the ability to maintain chronic threat, either internally through thoughts, or externally through modern life circumstances.

How thoughts alone maintain pathostasis

In the 1980s researcher Susan Nolen-Hoeksema, who would later become chair of psychology at Yale, began studying what she called "ruminative thinking.” She defined rumination as repetitive, passive thinking about one's problems, their causes, and their consequences. So basically when your thoughts circle the same worries over and over without resolving anything. Through decades of longitudinal studies following thousands of participants, she and other researchers documented that people who ruminated more developed a whole range of physical health problems at higher rates than those who didn't. Rumination reliably predicted future illness even after controlling for current health status. Something about the thinking pattern itself seemed to be driving disease.

In 2006 Brosschot and his colleagues explained how and why rumination is causing these outcomes. They realized that while a stressful event might last minutes or hours, rumination can keep the threat response activated for days, weeks, or years. Your body can't tell the difference between a threat that's currently happening and a threat you're just vividly imagining.

In 2014, Peggy Zoccola's team at Ohio University brought healthy young women into the lab and had them give a stressful speech. Then they randomly assigned half the women to ruminate about how the speech went, while the other half were distracted with neutral thoughts about sailing ships and grocery stores. The researchers drew blood samples and tracked inflammatory markers, and the women who ruminated showed sustained elevation in C-reactive protein, an inflammatory marker, that continued rising for at least an hour after the speech. While the markers of the women who were distracted returned to baseline.

A major meta-analysis published in Psychological Bulletin in 2016 synthesized the research across dozens of studies. The findings consistently showed that rumination and worry were associated with elevated blood pressure, elevated heart rate, elevated cortisol, and reduced heart rate variability. The same physiological patterns that show up in chronic disease. Thinking about a threat was producing the same body chemistry as facing one.

In 1997, Kubzansky and her colleagues published a landmark prospective study where they followed 1,759 men without heart disease for twenty years. They measured their worry levels at the start, and after controlling for cholesterol, blood pressure, smoking, and diabetes—all the standard cardiac risk factors—worry independently predicted who would have heart attacks. Men in the highest worry category had more than double the risk compared to those in the lowest category, and there was a clear dose-response relationship. Meaning the more they worried, the more heart attacks they had over a two decade period.

The second door

But rumination isn't the only way our cognitive brains keep pathostasis running. There's a second expression of override that looks completely different from the outside but produces the same chemistry underneath.

In the 1990s, Stanford researcher James Gross started studying what happens in the body when people suppress their emotional responses. He'd show participants films designed to provoke disgust or sadness, and tell some of them to watch the film without showing any reaction - to sit there looking calm no matter what they felt. And what he found was that the people who suppressed their reactions successfully on the outside, showed an increase in threat activation internally, whether the emotions they were suppressing were positive or negative. And this research has been replicated dozens of times, across cultures, ethnicities, ages and genders over 30+ years of research. Demonstrating that our unique human ability to change our physical response to its internal cues can turn on and maintain our threat response physiologically.

The health consequences for people who do this often have been well documented as well, and are what you'd expect given everything we've been tracking. Habitual suppressors show elevated C-reactive protein, the same inflammatory marker that Zoccola found in the ruminators. A one standard deviation increase in habitual suppression is associated with significantly elevated C-reactive protein, even after controlling for demographics, socioeconomic factors, and health behaviors.

And there's a whole population of people doing this as their default mode. Researchers identified a group they call "repressors" - people who score low on anxiety questionnaires but show high physiological arousal when measured. Meaning they report being fine, while their bodies show elevated threat response chemicals. In one study, researchers measured skin conductance and self-reported distress continuously, in real time, while participants watched emotional films. The repressors showed high skin conductance and low self-reported distress simultaneously, which means even in real time, these people thought they were calm as their bodies were physiologically activated.

Repressive coping accounts for up to 50% of certain populations, and the prevalence increases with age. This population shows a two-fold increased risk of death, heart attack, and cardiac events, even after controlling for other variables. And there are demonstrable links to coronary heart disease, hypertension, asthma, and cancer outcomes. Thirty years of research documenting that the people who say "I'm fine" while their chemistry says otherwise are getting sicker at dramatically higher rates.

So rumination keeps pathostasis running by generating more threat through thought. Suppression keeps it running by overriding the body's signals and pushing them down. Two completely different behaviors, same chemistry underneath.

Structural brain changes

In 2002, researchers at the National Centre for Biological Sciences published a foundational study in the Journal of Neuroscience demonstrating how chronic threat chemistry reshapes neural architecture. Vyas and colleagues subjected rats to chronic immobilization stress and then examined their brains at the cellular level. They found that in the hippocampus—the region that helps regulate memory, emotion, and the threat response itself—chronic threat chemistry caused dendritic atrophy. Meaning the neurons shrank and their branches retracted, and the structures they use to communicate with other neurons withered. These changes were visible under a microscope, and they correlated with impaired function. One of the functions of the hippocampus is to help stop the threat response, putting on the brakes and providing negative feedback to calm things down after a threat has passed. If you shrink the hippocampus, you weaken those brakes.

In the amygdala—the brain's threat detection center—the opposite happened. Neurons sprouted new dendrites, grew new branches, and formed more connections. The amygdala got stronger, more elaborate, and more connected to everything else. So the region responsible for triggering the alarm expanded its capacity while the region responsible for calming things down contracted.

The prefrontal cortex, which is the part of the brain responsible for executive control, and for consciously overriding automatic responses, shows similar changes. This is the part of the brain that can say "I know this feels dangerous but it's actually fine." Chronic threat chemistry causes dendritic retraction here too, weakening the very circuits that allow us to regulate our emotional responses. More recent research has revealed how this happens: microglia, the brain's immune cells, become overactive under chronic threat and start pruning synapses in the prefrontal cortex. A 2023 study showed that threat chemistry elevates complement C3, which tags synapses for elimination, and microglia then engulf them. The brain is literally dismantling its own regulatory capacity.

These changes persist even after the threat is gone. The brain has been architecturally reorganized to maintain a state of vigilance: stronger threat detection, weaker threat regulation, reduced capacity to override the alarm once it's triggered. And this reorganized brain is now more likely to ruminate, which creates a feedback loop. In 2011, J. Paul Hamilton's team at Stanford used functional MRI to examine the brains of people who ruminated frequently, and they could actually see the neuroplastic maps that were being built through these thought patterns. Just as Merzenich saw in the monkeys. The pathways for self-focused worry had been practiced and strengthened, making rumination the path of least resistance.

So let's bring it all together. Rumination maintains pathostatic chemistry even in the absence of actual threat—your thoughts alone keep the threat response firing. Repressors maintain it by overriding their body signals. Those chemicals physically remodel the brain: the hippocampus shrinks, the prefrontal cortex loses synapses, the amygdala grows and becomes hyperconnected. And this remodeled brain is now wired to regulate less effectively, which maintains pathostasis, which drives more remodeling. Each piece making this feedback loop harder and harder to interrupt. Researchers have mapped each of these connections across thousands of papers.

We now can see that the reason humans can create and maintain pathostatic conditions is because we evolved our cognitive abilities way too fast for them to integrate effectively with the 500 million year old threat detection system we share with all vertebrates. This is the engine fueling pathostasis. And it turns out, one field has been circling the answer for addressing it for over a century. Let's look at how close they got.