Transcription
So, I'm Chris Palmer. Um, I'm a psychiatrist, and I'm going to be talking with you about the metabolic theory of mental illness. I really want to thank Emily and all for inviting me. Greg, um, just a quick show of hands—how many people have heard me give this type of lecture before? Good, no, I'm really glad. So, I'm sorry to the people who've heard it, but that's what I usually—I usually assume people haven't heard it. Um, at some point, I'll give the advanced course, but today, we're just going to walk through it. So, I have no financial disclosures related to the work that I do, other than that I wrote a book and I get royalties for that book.
In case you didn't know, we have a mental health crisis. Prior to the pandemic, about 1 billion people on the planet suffered from a mental disorder. That represents about 13% of the world's population. Rates in the United States are even higher—one in five people has a mental disorder in any given year, and about 50% of Americans will meet criteria for a mental illness at some point during their lives. But, because of the pandemic, rates are even higher now. Mental disorders are one of the leading causes of disease burden and disability worldwide. Yes, access to care is sometimes problematic, but even for people getting the best care, our treatments often fail to work.
There's this curious thing: the rates of mental illness across a wide range of diagnostic categories have been increasing for decades. Rates of autism have quadrupled in the last 24 years; rates of ADHD have tripled since 2010. But these statistics are actually not the most recent; there's been an acceleration in the rates of diagnosis. In case you didn’t know, we’ve had an Adderall shortage the last few years. The rates of depression last year reached an all-time recorded high, both in terms of current prevalence and lifetime prevalence. This is a worldwide phenomenon. The rates of eating disorders in the United Kingdom, for example, have led to a more than fivefold increase in hospitalizations for eating disorders. These are the people who are too thin, who are starving themselves to death. You do not get admitted to the hospital for binge-eating disorder; you get admitted to the hospital for starvation. Overdose deaths in the United States are up over fivefold.
So, what are we doing about this crisis? Well, we have treatments—a plethora of medications, psychotherapies, hospitalizations, group therapies, electroconvulsive therapy, transcranial magnetic stimulation, ketamine, psychedelics, addiction recovery programs, and eating disorder programs. I am a Harvard Medical School psychiatrist; I've been there for almost 30 years. I want to say that these treatments, in fact, do work. They save millions of lives every year. I am not here to bash the treatments. But, unfortunately, even with the best treatment possible, they often don’t work. Treatment resistance is very common, affecting anywhere from 20% to 60% of people diagnosed with psychiatric disorders. Why such a ridiculously wide range? Because it really depends on the diagnosis. How long are we following people? What’s the follow-up methodology? Are we talking about lasting remission? If they get a remission for two months and then it comes back, is that treatment-resistant, or do we count that as a response because they responded for two months? The reality is, this is a crisis—a massive crisis taking a massive human toll. And it begs the question: why can’t we do better?
I would argue that we can't do better because our field has been unable to answer two fairly simple questions. The first question is: What exactly, Dr. Palmer, causes mental illness? I can't tell you how many times I've been asked this. If you ask the leading psychiatrists and neuroscientists in the world, they will tell you—if you find an honest one—they will say, "No one knows." It’s just too complicated. All we know are risk factors. We talk about the biopsychosocial model, so this is what you’re going to hear: it’s neurotransmitters, genetics, hormones, trauma, stress, childhood adversity, loneliness. Loneliness is the big thing for the Surgeon General right now. Tackling loneliness, they say, will tackle our mental health epidemic. But how do all of these things fit together? Well, they’ll say, "Dr. Palmer, it's just too complicated; no one knows."
The second common-sense question that our field has been unable to answer is: What exactly is mental illness? On the surface, this seems like a ridiculous, rhetorical question because we have the DSM-5-TR, which has all these household labels that you know: schizophrenia, depression, bipolar disorder, alcohol use disorder (a.k.a. alcoholism), anorexia nervosa, OCD. On the surface, most people assume these are valid disease constructs; they’re household words. But there are three problems with DSM constructs: heterogeneity, comorbidity, and a lack of validity.
Heterogeneity means that two people with the same diagnostic label can have wildly different symptoms from each other. One person with depression can be overeating, oversleeping, and have no suicidal thoughts. Another person can be undersleeping, undereating, losing weight, and plagued with suicidal thoughts. And I could go on with other symptoms, but they can have wildly different symptoms, sometimes polar opposite symptoms, and we give them the same label. Two people with autism—one could be a child whose parents painfully have to put him in a group home because they can’t safely care for him at home; he bangs his head against the wall until he bleeds, has seizures, and profound cognitive impairment. If you ask, "What’s wrong with this child?" they’ll say, "He’s autistic." And yet we see billionaire businessmen on the news every day with this exact same diagnostic label.
Comorbidity means that if you have one mental disorder, you’re more likely to have another. On average, population studies show that if you qualify for one mental disorder, there's a 50% chance you qualify for more than one. But if you're actually seeking treatment—mental health treatment—on average, people have three to four different diagnoses. So, if you're being treated for schizophrenia, there’s a good chance you also have OCD, maybe a substance use disorder. If you've got anorexia nervosa, you might also have OCD and a substance use disorder. Major depressive disorder? You might also have OCD and a substance use disorder. Now, on the surface, schizophrenia, depression, and anorexia nervosa are wildly different things. But if you look at real humans who have these disorders, they often overlap on two out of three of their brain disorders. How does that work?
This is nothing new. If you haven’t heard this before, this is not heresy. Tom Insel at the NIMH figured out decades ago that we’re never going to make progress in this field as long as we cling to these failed labels. Under his leadership, the NIMH developed the Research Domain Criteria, a transdiagnostic model for diagnosing mental disorders. There are also competing models, like the Hierarchical Taxonomy of Psychopathology. These models get complicated fast because there are hundreds of different mental health symptoms. Some of them can be considered normal and others abnormal. So, the field has just stuck with schizophrenia and depression. All of this says that the mental health field is a mess. And it’s tragic because people are dying. People are suffering. I’ll bet money that most, if not all of you, know someone—a nephew, niece, son, daughter, the neighbor’s kid, your coworker's kid. It's not funny.
The shocking thing is that several lines of research suggest this problem might actually be solvable. And all of these lines of evidence suggest that mental disorders share a common pathophysiology. In 2012, researchers like Leu examined internalizing and externalizing mental disorders and found that they seem to converge in some way, suggesting a common factor. In 2018, Caspi and Moffitt expanded on this model and looked at every label in the DSM, finding the same conclusion. Then, last year, Brandt et al. suggested that it's not just mental disorders—many physical health conditions also seem to share this common pathophysiology. These are mathematical models examining risk factors, genetics, and other factors. The models couldn't specify what the common pathway is, so many in the field think this can't be true. They argue it’s unimaginable that something could unify all mental disorders. Others strongly believe they know what it is. You’ve probably heard it before: trauma. Trauma causes all mental illness. Figures like Gabor Mate and others argue that trauma is the unifying cause of all mental illness. There’s a lot of controversy in the field.
What I have suggested is that it’s not trauma—though trauma can play a role for some people—but that mental disorders are metabolic disorders affecting the brain. And why would I say this? Because every known risk factor for mental illness can be tied directly to metabolism. Metabolic dysfunction in cells can explain why the brain exhibits the symptoms we see in people with mental disorders. This approach integrates biopsychosocial risk factors and stops the artificial separation between mental and physical health.
For those who haven't heard this, I want to share a quote from Albert Einstein: “Creating a new theory is not like destroying an old barn and erecting a skyscraper in its place. It is rather like climbing a mountain, gaining new and wider views, discovering unexpected connections between our starting point and its rich environment. The point from which we started out still exists and can be seen, although it appears smaller and forms a tiny part of our broad view gained by the mastery of the obstacles on our adventurous way up.” What Einstein meant is that his theory didn’t replace the theory of gravity. It integrated it. His theory unified the theory of gravity with other theories, like momentum. And that’s what I’m proposing in the mental health field. Trauma does play a role; neurotransmitters do play a role. This is a way to integrate it all.
Let me share a quick tour of some lines of evidence. There are actually numerous lines of evidence—I wrote a whole book and could probably have written ten more books on the topic. But let's start with correlations. Correlations don't equal causation, but strong bidirectional correlations make you consider the possibility of a common pathophysiology. Mental disorders share strong bidirectional relationships with disorders like obesity, diabetes, cardiovascular disease, and Alzheimer's—all of which people largely accept as metabolic. Bidirectional means that if I go to a mental health hospital, the patients there are more likely to have all of these disorders on the right side of the screen. But if I go to a Weight Watchers group, for example, and look at a group of obese people, they're more likely to have mental illness. Cardiovascular disease patients are more likely to have mental illness. People in nursing homes, likewise, have higher rates of mental illness.
Let me walk you through some statistics. If you have a chronic mental disorder, you're two to three times more likely to have obesity and diabetes. You're 50% to 350% more likely to have premature cardiovascular disease. Chronic depression? You’re two to three times more likely to develop Alzheimer's disease. If you have schizophrenia and live to the relatively young age of 66, you have a 20-fold, or 2,000%, increased risk of developing dementia on top of your schizophrenia. What about the other way around? If you have obesity, you’re 50% to 350% more likely to develop a range of mental disorders. Obesity increases the likelihood of depression, anxiety, personality disorders, psychotic disorders, and some—but not all—substance use disorders. Diabetics are two to three times more likely to develop a mental illness. Cardiovascular disease increases the risk three to fivefold. Alzheimer's patients? A study showed that 97% will have symptoms of mental illness—depression, anxiety, OCD, ruminations, insomnia. Forty to fifty percent will have hallucinations and delusions, which are the hallmark symptoms of schizophrenia. Consider this: if you have schizophrenia, you have a 20-fold increased risk of developing premature dementia. If you have dementia, you have a 50-fold increased risk of developing the hallmark symptoms of schizophrenia. We cannot talk about the root causes of any of these disorders without considering how they connect. And metabolism encompasses more than obesity, diabetes, and cardiovascular disease. You can be thin and metabolically unhealthy. You can be anorexic and metabolically unhealthy.
In fact, a large study showed that if you’re first diagnosed with a mental disorder, you're at greater risk of developing a range of physical health conditions—many of them metabolic. A recent cohort study of 170,000 adults, half with neuropsychiatric disorders and half as healthy controls, analyzed data from brain scans, blood work, and more. The conclusion? Poor body health, particularly of the metabolic, hepatic, and immune systems, was more marked as a manifestation of mental illness than any brain changes that could be measured or detected.
We’ve long known that people with chronic mental disorders die early deaths, losing anywhere from seven to 30 years of life, depending on their diagnosis. But in fact, a study published five years ago in The Lancet showed it’s not just depression, schizophrenia, and bipolar—it’s all of them. All DSM-5 labels are associated with premature mortality. On average, men with mental disorders lose 10 years of life, and women lose seven years of life. Now, everyone thinks that these deaths are due to suicide, that they’re "just crazy and killing themselves." Yes, suicide rates are higher, and yes, that plays a role, but the primary causes of death among the mentally ill are heart attacks and strokes—just seven to ten years younger than everyone else.
This has not gone unnoticed. The World Health Organization put together a task force, which reached three conclusions: factors related to the individual's behavior (the mentally ill overeat and under-exercise, or maybe it’s medication side effects), problems at the health system level (the mentally ill don’t go to their primary care doctors and take statins as they should), and wider societal issues (the mentally ill are more likely to be poor and unemployed, which correlates with early death). In reality, these explanations often blame the victim. We imply that the mentally ill are lazy slobs, that they overeat, don’t exercise, don’t work, and that’s why they are the way they are. We suggest that they’re just getting what they deserve, with choices stemming from a lack of willpower, laziness, and lack of discipline. But I am here to assert that these are actually symptoms of metabolic dysfunction. When you are metabolically compromised, you will be hungry when you shouldn’t be. You will be fatigued and lethargic in a way that’s hard to imagine. Your brain won’t work right. You won’t be able to hold a job, think straight, or remember things. You won’t have the motivation or executive function to show up at work on time every day.
This theory is not created out of thin air. It’s about integrating over a century of science. In the 1930s, researchers discovered abnormalities in lactate metabolism and glucose metabolism in people with bipolar disorder and schizophrenia. We've been doing neuroimaging studies for decades, with PET scans, SPECT scans, and functional MRI, which all measure brain metabolism. The diagnoses are widespread—depression, anxiety, PTSD, schizophrenia, bipolar, anorexia, and alcohol use disorder. So, Dr. Palmer, what’s the test for metabolic dysfunction? There isn’t one. There’s a wide range of tests that may be different for different people. But when we look at groups of patients with mental disorders compared to healthy controls, we see differences in lactate, glucose metabolism, ATP to ADP ratios, reactive oxygen species, Redux markers, levels of NAD to NADH, inflammatory cytokines, cortisol, brain insulin, and mitochondrial function. Now, you don’t need to understand the details, but know that metabolism is super complicated, and we don't have a single biomarker for metabolic dysfunction.
Some people struggle with this—they say, "Chris, come on. I see my patients in front of me; I put them on pills, and they clearly overeat, gain weight, and smoke like a chimney. They’re not exercising; that’s why they’re dying." This study suggests it might not be so simple. They followed children from the ages of 1 to 24. Those with the highest levels of insulin resistance beginning at age 9 had a fivefold increased risk of developing a psychosis-risk mental state by the time they turned 24. They were three times more likely to be diagnosed with schizophrenia or bipolar disorder. The kids who gained the most weight around puberty were four times more likely to develop clinical depression by 24. People like to think in linear terms: bad food makes you sick, and then you're metabolically unhealthy. But think of it as a feedback loop—a cycle between what we call mental health and metabolic health.
So, metabolism—what does that mean? People often say, "It’s burning calories, Dr. Palmer." Or, "It’s metabolic syndrome, with blood pressure, glucose, insulin resistance." Both are true but insufficient. Metabolism is the process that all living organisms use to convert food into energy and building blocks to maintain or grow cells. It also involves waste management. Metabolism is fundamental to the definition of organisms. In fact, the cessation of metabolism defines death; there are no exceptions. If an organism can’t do metabolism on its own, like a virus, it’s not considered a living organism. So, if I’m using a broad definition of metabolism, that may seem too all-encompassing to be helpful. But let’s keep going.
What controls metabolism? We’re led to tiny things in our cells called mitochondria. In biology class, you may have heard that mitochondria are the "powerhouse" of the cell. Cutting-edge research over the last 25 years has completely decimated this simplistic definition. Yes, they are powerhouses, but they are so much more. They’re the master regulators of metabolism. There are hundreds or thousands of them in each cell; they form communities, move around, fuse, and bud off from each other. They’re involved in numerous cellular processes. If you start diving into mitochondrial science, it gets overwhelming fast. They have different cell-dependent phenotypes, features, activities, functions, and behaviors. They’re a universe within the cell. And we know enough to say that they are the common pathway to neuropsychiatric disorders.
Mitochondria control the production, synthesis, and release of neurotransmitters like GABA, glutamate, serotonin, dopamine, and others. But these neurotransmitters also impact mitochondrial function. Mitochondria are the primary creators of signals that control epigenetics from the cell nucleus. But genes from the cell nucleus also influence them. They regulate inflammation, turning it on and off, though inflammation and inflammatory biomarkers affect their function. The gut microbiome, diet, and exercise affect them too. Manipulating mitochondria in brain cells can make animals obese or thin. Mitochondria control whether you have the energy to exercise, but exercise improves their function if done right. Trauma, stress, drugs, alcohol, sleep, and hormones all affect mitochondrial function. Mitochondria even control the first steps in synthesizing hormones like cortisol, estrogen, testosterone, and progesterone, known to play roles in both mental and metabolic health.
When I first published this, I thought everyone would laugh at me. Fortunately, others are joining in. Some researchers even cited me, saying, "Yes, it's mitochondria." They are the way to unify all biopsychosocial risk factors. Stress and trauma directly influence metabolism through anticipatory responses, but if prolonged, they result in maladaptive outcomes converging on mitochondrial dysfunction. Again, this is not new. It integrates a century of science. In the 1800s, diabetes, a metabolic disorder, was observed as common in families with a history of insanity. Within 10 years of discovering insulin, we were injecting it into the mentally ill as a treatment. We observed abnormalities in lactate and glucose metabolism in bipolar and schizophrenia patients in the 1930s. The mitochondrial theory of autism was proposed in 1985. Imaging studies, like PET and SPECT scans, support the mitochondrial theory of bipolar disorder proposed in 2000. Since then, there has been an exponential increase in research from neuroscientists, psychiatrists, and biotech companies.
A new theory has to open up entirely new doors. It has to have predictive value. Time will tell if it does. But in the last five years, I’ve been able to predict a lot of research I didn’t even know existed, just by asking, "If my theory is true, could marijuana affect mitochondria and cause mitochondrial dysfunction?" I Googled it, and sure enough, it was already published. That’s why people get psychotic on marijuana. I’ve gone through hundreds of examples. But as a psychiatrist, I want to know what we can do for patients right now. This theory suggests we need a new model of healthcare—one that integrates metabolic health, mental health, and longevity. Some may say we’re already doing this, but the reality is, we don’t offer integrated healthcare. If you go to a mental health professional, there’s a good chance they’ll give you pills that make you overweight and diabetic, increasing your risk for cardiovascular disease. This isn’t conspiracy theory; the risks are listed right on the package inserts.
What can we do now? First, I don’t want to throw the baby out with the bathwater. Treatments work. They really save lives. We need to keep using them for now. But the things we rarely discuss in mental health are diet, nutrition, vitamins, supplements, and exercise. And regarding medications, almost all, if not all, psychotropic medications can be tied directly or indirectly to metabolism and mitochondria. One group of researchers looked at all the animal models of Prozac, which is not due to a serotonin imbalance in the brain. In fact, 20 years of evidence disproves this. These researchers found that Prozac increases brain energy metabolism through oxidative phosphorylation and mitochondrial changes. Similarly, Celexa and Lexapro work by increasing mitochondrial energetics. Lithium impacts the insulin signaling pathway, converging on mitochondria. But while some medications enhance brain energy metabolism, others cause metabolic harm, like weight gain and diabetes.
So, can a metabolic treatment play a role in treating mental illness? Let’s consider some metabolic medications. Statins, calcium channel blockers, metformin, and SGTL-2 inhibitors, as well as GLP-1 receptor agonists, are all diabetes drugs that also help with depression and substance use disorders. Diet can play a role, too, with interventions personalized for different patients. For example, the ketogenic diet has been around since the 1800s for type 2 diabetes and has shown efficacy in epilepsy, which we treat in psychiatry. How does a diet stop seizures? It’s not just a "healthy diet"; it’s a metabolic intervention for people who are metabolically impaired. The ketogenic diet improves mitochondrial function, a process we call mitophagy. When people use it for seizure control, they often go off the diet after a few years and maintain the benefits. Does it work? Yes. NIDA director Nora Volkow has shown that the ketogenic diet improves brain energy metabolism, reducing withdrawal symptoms, cravings, and inflammation in the brain. In a study of bipolar and schizophrenia patients, 43% achieved clinical remission. A study of treatment-resistant bipolar disorder and schizophrenia patients showed that those on the ketogenic diet lost weight and experienced improved brain metabolism.
I’m now honored to be running a metabolic mental health program at McLean Hospital. But this is a global effort. We have research studies at Harvard, Oxford, Johns Hopkins, and Stanford, as well as in various places around the world. Thank you.