From Symptom Control to System Repair in Psychiatry

Imagine an office building where the lights flicker every afternoon, the elevators run slowly, and the ventilation struggles on hot days. Facilities management responds by replacing bulbs, resetting the elevator controller, and adding fans. Each fix works for a while. Nobody has yet checked the building's electrical supply, which is undersized and degrading. The complaints are real, the responses are reasonable, and the underlying problem is untouched. This is a fair description of what much of psychiatry has been doing well and doing incompletely at the same time.

The mechanism is worth understanding properly, because it is not a metaphor. Brain tissue is metabolically extraordinary: roughly two percent of body weight consuming around twenty percent of resting energy. That energy is spent on the constant work of restoring electrical gradients across neuronal membranes, packaging and releasing neurotransmitters, and rebuilding synapses. The currency is ATP, and the factories are mitochondria — small organelles inside every cell that convert fuel into usable energy. The brain's default fuel is glucose, but it can also run on ketone bodies, molecules produced by the liver when carbohydrate availability falls. The ability to switch cleanly between these fuels according to what is available is called metabolic flexibility. When that flexibility is lost, neurons do not simply slow down evenly. Energy-hungry functions degrade first: sustained attention, emotional regulation, the plasticity that allows learning and recovery from stress.

Three dimensions of this deserve a closer look. The first is mitochondrial function itself. Mitochondria are unusually sensitive to their environment, responding to oxidative stress, inflammatory signals, and fluctuations in fuel supply. Studies across major depressive disorder and bipolar spectrum illness have repeatedly found markers of altered mitochondrial dynamics and reduced energy output. Whether this is cause, consequence, or both is not fully settled, but mitochondria function as a reasonable barometer of how much reserve a neural network has when demand rises. The second is insulin resistance. Most people associate it with blood sugar and type 2 diabetes, yet insulin also acts as a signaling molecule in the brain, influencing how neurons take up fuel and how readily synapses remodel. When that signaling blunts, the orchestration of mood, motivation, and cognition suffers. The third is the loop between inflammation and metabolism. Pro-inflammatory cytokines impair mitochondrial performance; impaired mitochondria generate more oxidative stress and more inflammatory signaling. Chronic stress, poor sleep, infection, or metabolic syndrome can push a person into this loop, and the loop is self-sustaining.

In ordinary working life this rarely announces itself dramatically. It looks like a person whose medication helped considerably — the acute crisis passed, the intrusive thoughts quietened — but who still cannot get through a two o'clock meeting without their thinking going gluey. It looks like partial remission: the symptom scale improved, the life did not fully return. Clinicians call the leftovers residual symptoms, and fatigue, apathy, and cognitive slowing dominate that list. These are precisely the symptoms you would predict from a system with adequate signaling but constrained energy.

Several misreadings follow easily here, and they are worth heading off. The first is treating this as an argument against medication or therapy. It is not. Established treatments remain essential, particularly for stabilization and where risk is high, and no one should alter psychiatric medication without their prescriber. The metabolic view layers on top of conventional care rather than replacing it. The second misreading is assuming the neurotransmitter story was simply wrong. It was not wrong, it was downstream — describing the signaling that energy metabolism makes possible. The third, and the most common, is collapsing all of this into a diet recommendation. Ketogenic approaches are genuinely mechanistically interesting, and the evidence in drug-resistant epilepsy is strong, but the psychiatric evidence consists largely of pilot studies and small trials. That is promising, not proven. Such approaches also carry real risks — nutritional deficiency, electrolyte disturbance, interactions with medication and with existing medical conditions — and are unsuitable for several groups, including people with certain metabolic disorders, a history of eating disorders, or pregnancy. They belong under clinical supervision, not self-experimentation.

Here is the observation that tends to surprise people most. In many cases the brain is not short of fuel. Blood glucose may be perfectly adequate, even elevated, while neurons behave as though they are starving. The problem is not supply but uptake and conversion — abundance the machinery cannot use. Which means that feeling depleted is not always evidence of doing too much. Sometimes it is evidence of a system that has lost the ability to convert what it already has.