The Invisible Famine

So now it’s time to come back to GLP1s and see what they’re telling us that our siloed approach is missing. Because they’re telling us a lot. The INFORM survey of 550 people found 62% had constant food thoughts before semaglutide, dropping to 16% after. A Cleveland Clinic doctor describes it as "a light switch going off, where all of a sudden they have some freedom in their brain and relief from constant signals." Patients use phrases like "the food noise was gone" and "mental quiet."

But no one seems to have a coherent answer to WHY there is so much food noise in a population that is by definition getting enough food. One hypothesis is that processed food isn’t triggering satiety the way whole foods are, but that accounts for an extra 500 calories of consumption per day, not constant food noise. And no one eating whole food optimized diets is reporting their food noise going away. If the processed foods were the culprit, eating whole foods should be reducing food noise too, and no one’s reporting that, so what else could it be? The other explanations are environmental and emotional triggers, hormonal imbalances, restrictive dieting, not eating enough, not sleeping enough, dehydration, lack of mindfulness. But do any of those account for the fact that people are reporting having had constant food noise for years? And if this is such a wide spread phenomenon, that would require believing either: everyone has the same one of these as a cause, OR that ALL of these explanations happened to produce food noise in humans? That not optimizing our inputs of sleep and food and our environment universally created the desire to eat?

Additionally, GLP1s are affecting many many things in addition to weight loss. If it were a food phenomenon, then this list wouldn’t make sense. Let’s look at the current landscape of what is being treated using this medication:

In addition to reduced food noise GLP1s are FDA approved for treating:

Type 2 diabetes Weight loss Cardiovascular disease risk reduction (heart attack, stroke) — and notably, this benefit was shown to be INDEPENDENT of weight loss Chronic kidney disease (in diabetic patients) * MASH / fatty liver disease

There is strong clinical data showing:

Heart failure (European approval, FDA pending)
Alcohol reduction / alcohol use disorder
Knee osteoarthritis pain
Asthma symptom control (60,000 patient study)
Fibromyalgia (pain, fatigue, opioid reduction)
Reduced risk of Alzheimer's (70% reduced risk vs insulin)
Reduced risk of Parkinson's (32% lower in meta-analysis of 450,000 people)
Reduced gambling and compulsive shopping (behavioral)
PCOS improvement
Improved fertility

Early/emerging treatment targets:

IBS pain
CFS/ME (trials underway)
Chronic pain broadly
Cancer risk reduction (some data, also some concerns)
* Sleep apnea

And the explanations for why it’s doing all of this are equally far reaching. Researchers are calling this a 'pipeline in a pill' and scrambling to explain each benefit through the lens of their own specialty. The gut researchers see gut mechanisms. The neuroinflammation researchers see neuroinflammation. The reward researchers see reward pathways. Everyone is attempting their own explanation. But given everything we’ve seen as we’ve walked through all of the medical data, let’s step back and see if there’s anything we can find that would explain it all through a single mechanism rather than 10 separate ones for each downstream effect.

So what are we seeing? We’re seeing something that affects seeking behaviors: food, alcohol, shopping, gambling. And disease progression and outcomes. And also things like mood, sleep, and pain. What could possibly explain all of these disparate seeming expressions?

Mechanism

The primary way that GLP1s work in the brain is this:

Your hypothalamus has two sets of neurons that drive seeking behavior. On one side you have AgRP/NPY neurons which are your need neurons, your hunger neurons. When they fire, they drive seeking behavior. They say: you need something, go find it. On the other side you have POMC neurons which are your satiety neurons. When they fire, they say: you have enough, stop seeking.

What GLP1s do in the brain is signal the POMC neurons, which subsequently quiets the AgRP neurons. And this area of the brain, the need center of the hypothalamus, controls more than just hunger signaling, it controls thirst signaling too. And they’re finding that people on GLP1s aren’t drinking enough water, so much so that they are telling people to set alarms on their phones to remember to drink more water because the need signal has been effectively turned off or down.

So in a population of millions of people that are getting enough food, and in a society where we have more than enough water to drink, why would turning off a signal saying “you need more” stop people from eating when they clearly already had enough to eat? Why were they getting that signal in the first place? Why would it stop people from gambling and shopping and drinking alcohol? What people are describing is that it’s like an alarm that has been going off in their head for as long as they can remember has been finally turning off. What could account for that?

The invisible need

There are two mice studies that together paint this picture better than any explanation we can walk through so let’s look at those first.

The first study took mice and isolated them right after they finished weaning through adolescence. These mice had a perfect diet, and every other need met, and from isolation alone they showed:

Type 2 diabetes
Overeating
Less sleeping
Inhibited social drive
Increased pain sensitivity
Less lean tissue and more body fat (they didn’t weigh more, just had less muscle)

Social isolation alone caused metabolic disease, not correlated with, caused. It caused higher pain sensitivity, more compulsive eating, and sleep issues. These are the exact same markers you see in chronically underfed or starving mice. These mice had every single other need met and isolation alone caused this whole clinical picture.

What our brains expect

Humans evolved for millions of years to live surrounded by dozens of people. The way apes currently live is a good analogue for the social world our brains expect. 30-80 people who are always available. Which means that when your brain sent a signal telling you to go get social connection, it was pretty much always available. You didn't have to "make time for connection" or "self care", those things were baked into the world we lived in. The ambient social environment was richly connected and the constant social signal and response was like air; it was the fabric of your reality every moment of your life.

In our modern world we see rich social connection as a nice to have. But in 2023 the U.S. Surgeon General put out an advisory saying social connection is "a fundamental human need, [and is] as essential to survival as food [and] water." The advisory went on to say that the impact on death risk "of being socially disconnected is similar to that caused by smoking up to 15 cigarettes a day, and even greater than that associated with obesity and physical inactivity", and calling loneliness and isolation a public health crisis.

A meta analysis that same year published in Nature, one of the most prestigious journals in medicine, looked across 90 studies including over 2 million people, and found that social isolation and loneliness increased overall death rates, and specifically heart disease and cancer related death rates. The effect size rivaling that of smoking.

Harvard neuroscientist Dulac took this idea a step further and actually studied loneliness and social need to find out where it lives in our brains, to see if calling it a basic need was true. Was it closer to an emotion like sadness? Or more akin to something like our need for food or sleep? And what she found was that it does not reside in the part of the brain that controls emotion, but that it was in fact located in the hypothalamus, which is the part of the brain that controls our basic biological human needs like hunger and thirst. It uses the exact same neural architecture. The way she framed it was to think of it like a bean counter. Your brain tracks your social input the same way it tracks your caloric input. When social input drops below threshold, the need neurons fire telling your brain “you need more”, and drive social seeking behavior. When social input reaches satiety, the satiety neurons fire and inhibit the need neurons. The same way it does for hunger or thirst.

And when we have a need signal that is being continuously unmet, the research shows this activates and keeps on the threat response. When you’re chronically hungry your cortisol levels elevate and remain high. In rat studies they’ve shown that even at 75% of the quantity of their daily intake of food their equivalent of cortisol was chronically elevated. And they’ve also studied this for loneliness. When your social needs are unmet, your HPA axis remains chronically activated. Chronically under met needs chronically elevate pathostatic chemicals. Which makes sense given that not getting enough to eat IS an existential threat. And evolutionarily, being socially disconnected was too. It would lead to death. It’s located in the same region of the brain that controls your pathostatic chemicals for a reason: because having a chronically under met need is a physiological threat.

In the 1960s, researchers studied a small Italian-American town called Roseto, Pennsylvania. The residents fried everything in lard, smoked cigars, drank wine regularly, and worked in hazardous slate quarries. By every metric medicine uses to predict heart disease, they should have been dropping dead. But what they found was that they had virtually no heart attacks in men aged 55-64, normally the highest risk group. The only thing that was different about Roseto was its social structure. The town had been settled by immigrants from a single Italian village, and they had recreated that village's tight-knit multigenerational social world almost exactly; three generations in one house, elderly central to daily life, dense interconnected relationships as the background fabric of existence.

Intermittent reinforcement

In the 1930s a researcher named B.F. Skinner was running low on food pellets one weekend and didn't want to go make more. So he set his apparatus to only reward the rats in his experiment sometimes instead of every time, which led to one of the biggest discoveries in learning and behavioral research. Skinner had been studying how rewards shape behavior - when a rat presses a lever, they get food. But when he came back Monday after his lazy weekend adjustment, the rats were pressing the lever like maniacs, far more obsessively than when they got food every single time.

When a behavior is rewarded unpredictably, the brain never knows when the next reward might come, so it keeps trying, the uncertainty itself drives persistence and compulsion. If you never know when the next meal will come it makes sense to keep checking. And it's not just rewards that work this way. The same mechanism applies to anything unpredictable, including threats. When danger appears sometimes but not always, the brain can't afford to let its guard down. It stays vigilant and keeps scanning for the threat, keeping the threat detection system running.

Because we evolved to be in tribes where you had dozens of adults and other children around you all the time, we evolved to have attunement pretty universally available. Which means that children’s bodies expect this. In our modern society where all you have is one or two busy, overwhelmed adults around you most of the time, what you get is textbook intermittent reinforcement of a basic human need. They need connection, they reach out, sometimes it’s there, sometimes it’s not, and sometimes it comes with threat attached. And over time through practice and repetition, those behaviors and feelings get wired right alongside connection.

Pairing threat with connection

In 1975 a researcher named Tronick at UMass in Boston studied what happens when a mother who is normally very responsive to her baby emotionally stays in front of them but goes completely expressionless for 3 minutes. What they saw was that there was a predictable cascade of behaviors that the infants in the study would enact. First, they would solicit reconnection, they would smile, coo, reach, point, anything they could think of to get their mother to reengage with them. When that didn't work they would escalate; they would start being louder, more insistent, exaggerate their expressions etc. When that didn't work they would show distress, they would frown, flail their limbs, show confusion or agitation, and eventually they would start crying. When all of this failed to work they would then withdraw. The baby would go still, start sucking on their fingers or thumbs to self sooth. Their facial expression would look withdrawn and hopeless. And this study has been replicated hundreds of times across many different cultures. And even more telling, they have measured cortisol levels and have seen a distinct elevation starting at phase three. This happens in just three minutes of social withdrawal. Three minutes to go through this entire cycle of distress. And they see these same behavioral patterns in infants of depressed mothers. Those mothers are the still face experiment all the time, and so they see them cycling through these patterns over and over again. The effects of adults looking down at their phones while caring for their infants are likely staggering.

If we next look at a more extreme version of social isolation, children raised in orphanages from birth where they were fed and changed but otherwise ignored, they also have a very clear pattern of dysfunction. These children show stunted brain development, physical growth delays, behavioral problems, autism-like behaviors like rocking, stimming, and hand flapping, and psychological issues like depression, anxiety, disruptive behavior, and adhd. The longer they were in the institution before adoption, the more these patterns stayed stuck throughout their lives. Which shows that the behaviors and responses to social connection or the lack thereof laid down in childhood persist.

Remember the ACEs survey from earlier in the book? With its ten questions that predict disease better than anything medicine routinely measures? Looking at those questions again, every single one is measuring social environment failure. With a dose response curve showing the severity of social starvation during development predicts how deeply pathostatic wiring gets laid down. Being in an environment with other animals, our parents, who are by design of modern life, intermittently available and likely overwhelmed themselves, as our primary social connection, leads to wiring in threat and compulsion alongside connection. Which means that even when we do have people in our lives, that social connection isn’t feeding us the way it could if it weren’t also coded in as a threat. And the cognitive loops we described in the last chapter, rumination and suppression, are how we try to manage this constant threat state internally. If you think about the content of rumination and suppression, it is always based around safety needs and the threat we are experiencing alongside them, and most of the time, in our modern world, that need is social.

The cognitive revolution gave us the ability to think differently than any other animal before us, and we used this ability to optimize for food security, and shelter, and water. For comfort, entertainment, and ease of life. We have objectively never been more safe, more secure, and more comfortable. And we have never been more chronically ill. Each innovation has stripped away more ambient social fabric, replacing it with longer commutes, smaller households, more screens, less unstructured time with other humans. And each generation raised with less social infrastructure raises the next generation with even less capacity to provide it. Zoo animals show the exact same disease and dysfunctional behavioral profile as humans, where wild animals do not. Zoo animals are fed perfectly optimized diets; if this were about the food, zoo animals should be the healthiest animals on the planet.

And this is why we see chronic illness rates on the rise. 60% of US adults have at least one chronic condition and 40% have multiple, with 90% of the annual healthcare spending going to chronic disease. And those numbers are only including the diseases that medicine considers "real". If you included the whole swath of diseases, the actual full picture of chronic suffering, adding things like fibromyalgia and chronic fatigue and IBS, that number gets closer to 80 or 90%. Even in children the rate of chronic disease has nearly doubled from 23% in 1999 to more than 40% today. We see the rates of chronic disease rising across all age groups, and nobody seems to know why.

Why GLP1s work so widely

When researchers gave mice GLP1s who were raised in a rich social environment, it actually decreased their social drive. Which makes sense right? You are telling the need center, we have enough, so they stop seeking, just as we are seeing with food. But when they gave GLP-1s to the socially isolated mice from the study above, it actually recovered their social drive. Because it reduced their threat and took them out of a deep need state they had been trapped in for most of their existence. It eased their dysfunction, allowing their natural instincts to surface.

To understand this better lets look at what we see in our modern world. If you picture a society where most people’s brains are screaming at them “you need more”, and these needs can be confused with each other as evidenced by GLP1 users not drinking enough water, and this need state is triggering pathostatic chemicals what would you expect to see? What do we see societally?

We see widespread compulsive eating. Compulsive drinking. Compulsive shopping. Compulsive scrolling. Workaholism. Political tribalism: seeking belonging and safety through ideological identity. Social media addiction: seeking connection through a channel that can't actually deliver it. Rising rates of every chronic disease. Rising rates of anxiety and depression. Rising rates of addiction across every category.

And then a drug comes along that flips that one switch; it tells the brain “you have enough” and we see people’s compulsions ease, their disease markers improve, and we see things like inflammation, anxiety, and obesity go down. Because GLP1s are doing two things. They are telling us we have enough, we can finally stop seeking. And they are turning off or down our pathostatic chemistry.

Researchers are seeing lower reward seeking and saying GLP1s act on reward circuits; but if you think about it through this lens, if you are starving and you eat food you get a huge reward. If you are full and you eat a meal you get far less reward chemistry. Taking the GLP1s is telling the brain “you have enough” which means the rewards are naturally lower when they consume or do addictive behaviors. These people didn’t have broken reward pathway signaling, they were just never getting the you’re good here signal to turn off the seeking.

Our cognition allowed us to optimize our environment for safety, and to become the most powerful animal on the planet. And in so doing we created a world wide famine that led to ever increasing rates of dysfunction and disease. GLP1s get us part of the way toward addressing the downstream consequences. But remember from the last chapter how when our brains are stuck in a chronic threat state they change structurally? Which means often, when we’ve spent years in a chronic deficit, just turning off the signal isn’t enough. In order to understand the full landscape and what we can do about it, let’s look at what we know about how our brains change, and what we can do about it.