Mental Illness As A Metabolic Condition: the brain energy theory

Much of the information in this post comes from a talk by Dr Christopher Palmer at a Harvard Medical School and McLean Hospital conference I attended in Boston in 2024.

Psychiatry has a long list of risk factors for mental illness and very little in the way of a mechanism that ties them together. Genetics, neurotransmitters, hormones, trauma, adverse childhood experiences, loneliness, poverty, sleep loss, diet and substance use all raise risk. The biopsychosocial model holds these side by side without explaining how they converge to produce symptoms. That gap has clinical consequences. Treatment resistance affects somewhere between significant number of people receiving psychiatric care and chronic mental illness remains associated with a substantially shortened life expectancy, driven largely by cardiovascular disease rather than by suicide.

The Proposition

The brain energy theory, articulated by Dr Christopher Palmer of McLean Hospital and Harvard Medical School, asserts that mental disorders are metabolic disorders of the brain. Psychological and social factors retain their full weight under this model. Each of them exerts its influence through one shared biological pathway and that pathway is cellular metabolism.

Metabolism in this context refers to the processes by which cells convert fuel into energy or into building blocks, maintain and repair themselves and clear waste products. It is a far broader idea than weight, blood glucose or metabolic syndrome. Mitochondria govern much of this work. Beyond generating ATP through oxidative phosphorylation, they participate in calcium regulation, reactive oxygen species signalling, steroid hormone synthesis including cortisol, inflammatory signalling and cell death. Monzel and colleagues have described them as a family of organelles with distinct phenotypes, activities and behaviours, which is a considerable expansion on the powerhouse metaphor most of us learned at school.

What the Theory Integrates

The first thing it accommodates is the mess of diagnoses. Psychiatric categories overlap heavily, comorbidity is the norm rather than the exception and successive research and practice reveals diagnostic criteria carry wide heterogeneity. Lahey proposed a general factor underlying internalising and externalising disorders, Caspi and Moffitt described a p factor common to all mental disorders and Brandt and colleagues have since extended this to some physical illnesses. A theory featuring cellular energy can accommodate that picture. Impaired mitochondrial function produces different symptoms depending on which brain regions and circuits are affected, which is what a shared pathway with variable expression should look like.

The second is the relationship between mental illness and metabolic disease. Obesity, type 2 diabetes, cardiovascular disease and Alzheimer's disease cluster with psychiatric conditions in both directions and across the lifespan. The temporal ordering is the interesting part. Metabolic disturbance frequently precedes serious mental illness rather than following it. Perry and colleagues found that children with the highest insulin levels from age nine were markedly more likely in early adulthood to have a psychosis at-risk mental state or an established psychotic disorder. Pillinger's work has shown glucose dysregulation at first presentation of psychosis, before treatment and before any diagnosis of diabetes. Tian and colleagues, comparing over eighty-five thousand adults with neuropsychiatric disorders against healthy controls, reported that poor metabolic, hepatic and immune health was a more marked manifestation of mental illness than brain changes were. This is difficult to explain as medication effect or lifestyle consequence alone.

The third is the mechanism of the treatments we already use. Large-scale gene expression studies of fluoxetine across 27 mammalian brain regions point to increased oxidative phosphorylation and mitochondrial change as a shared mechanism of action. Analysis of citalopram and escitalopram responders implicates acylcarnitine metabolism and beta-oxidation is key in who responds and who doesn’t. Lithium acts on insulin signalling pathways that terminate in glucose transport, mitochondrial respiration and ATP production. Statins, calcium channel blockers and metformin are each associated with reduced psychiatric hospitalisation and self-harm in bipolar disorder and schizophrenia. Metabolism appears to be part of how all of these agents work, despite none of them having been developed with that mechanism in mind.

The fourth is the boundary with neurology. Anticonvulsants developed for epilepsy have been used as mood stabilisers for decades. The ketogenic diet was developed in the 1920s as an epilepsy treatment before showing signal in psychiatric populations. A neurological condition and a psychiatric one respond to the same interventions, which is difficult to explain if the two are mechanistically unrelated. Alzheimer's disease points the same way. Brain glucose hypometabolism has been demonstrated and correlates directly with plaques, tangles and clinical symptoms, mitochondrial dysfunction is well documented and the risk factor list reads as a metabolic one, including diabetes, midlife obesity, cardiovascular disease, inactivity and poor sleep. The label "type 3 diabetes" emerged from exactly this observation. Depression and schizophrenia both substantially increase dementia risk. Almost everyone who develops Alzheimer's disease experiences psychiatric symptoms along the way. Psychiatric, neurological and neurodegenerative conditions look less like separate territories under this model and more like different expressions of impaired cellular energy in different regions at different stages of life.

Why this Matters Clinically

A theory earns its place by generating testable predictions and by changing what clinicians attend to. This one predicts that interventions targeting cellular energy should improve psychiatric symptoms, which is now being tested internationally in psychiatric populations. It also reframes metabolic health as core business in mental health care rather than as a side effect to be monitored. Fasting insulin, waist circumference, sleep, physical activity and dietary quality become clinically relevant to the presenting complaint itself.

This work belongs to a growing international field. Dr Christopher Palmer has done more than anyone to articulate the theory and bring it to a clinical audience through his book Brain Energy and the Metabolic and Mental Health Program at McLean Hospital. Dr Iain Campbell, Baszucki Research Fellow in Metabolic Psychiatry at the University of Edinburgh, has driven the bipolar research programme forward while living with bipolar disorder type II himself. His work on insulin signalling, metabolic plasticity and the evolutionary origins of bipolar disorder has shaped how the field thinks about mechanism. Professor Zoltan Sarnyai at James Cook University has built much of the preclinical foundation. The Baszucki Brain Research Fund and Metabolic Mind have moved this from a fringe idea toward a testable clinical science.

The strength of this framework is how much it explains that existing models leave unexplained. It gives a coherent account of why psychiatric and metabolic illness travel together, why psychiatric, neurological and neurodegenerative conditions share findings and respond to shared treatments, why diagnostic categories overlap so heavily and why medications developed for quite different purposes converge on the same cellular machinery. Controlled trials now under way internationally will refine the detail. What it already offers clinicians is a mechanism to work with in place of a list of risk factors to manage.

At Metabolic Psychology we work from this model every day. It shapes what we assess, what we measure and how we plan treatment, alongside the psychological and emotional work that has always been part of good mental health care. If you would like to discuss whether a metabolic approach suits your situation, you can book an appointment through our client portal.