A patient arrives at the office in full hypomania. Speech pressured, sleep gone for four days, judgment unraveling. In a conventional pathway, the next stop is an emergency room, a psychiatric hold, and an intramuscular injection of haloperidol or olanzapine. The mania quiets, eventually. The patient leaves the hospital sedated, often shamed, and with a brain that has been chemically restrained but not actually treated.

We don't run that pathway anymore.

Over the last several years, the literature on bipolar disorder has moved decisively toward an inflammatory and metabolic model. The diagnosis was historically framed as a disorder of mood, dopamine, and serotonin. The current picture is broader. Bipolar brains show elevated TNF-α, IL-1β, and IL-6 in serum, mitochondrial DNA damage, impaired lactate flux, and signatures of oxidative stress that look more like an autoimmune neuroinflammatory disease than a classical mood disorder. A 2024 review in Frontiers in Psychiatry describes the convergence plainly: bipolar disorder is, at its biological core, a disease of neuroinflammation layered on top of energy failure in high-demand neurons and glia.

That biology has clinical consequences conventional psychiatry rarely names.

What's Actually Failing in a Manic Brain

The prefrontal cortex is the body's regulatory governor. It restrains the limbic system, monitors errors through anterior cingulate circuits, and tunes the salience network so the patient can distinguish a small problem from a catastrophe. When metabolic substrate fails and neuroinflammatory cytokines saturate that circuitry, the frontal governor goes offline. What remains is limbic dominance, the kind of System 1 thinking Daniel Kahneman described, untethered from the slower, error-correcting top-down network that normally checks it. Resting-state fMRI work published in BMC Psychiatry in 2024 confirmed what bedside neurology has long suggested. Bipolar patients show reduced default mode network connectivity to medial prefrontal cortex and disrupted anticorrelation between the salience network and the executive control network.

In plain language, this is brain failure. The same framework we apply to addiction and to the collapse of error correction in substance use applies here. The behavior we call mania is the visible signature of a frontal network that has lost its energy supply and is drowning in inflammatory load.

If that is what is breaking, sedation is not treatment. It is suppression.

What Acute Treatment Should Actually Do

Our acute protocol targets the substrate. A patient in a hypomanic or manic crisis comes into the clinic and receives intravenous neuronutrients: alpha lipoic acid, a set of free amino acids, methylcobalamin (B12), thiamine, and magnesium. Layered with those are NAD+ to restore mitochondrial NAD/NADH ratios, glutathione to address oxidative stress in real time, and in selected patients, mesenchymal stem cell-derived exosomes for their anti-inflammatory and neuromodulatory cargo. Each of these has solid mechanistic ground under it. The 2024 literature on glutathione as a psychiatric adjunct, the work on MSC exosomes as immune modulators capable of crossing the blood-brain barrier, and the energetic literature on NAD+ in mitochondrial rescue all converge on the same logic.

The clinical part is what surprises people. Patients who would have been admitted and tranquilized for days or weeks instead come back down within an hour or two and are often able to resume normal activity in a day or two. Sleep returns. Speech slows. Judgment recovers. The behavioral noise quiets not because we have sedated the cortex but because we have restored fuel and tamped down the inflammatory fire destabilizing it.

The same logic runs in the opposite direction. A patient in deep bipolar depression, on the conventional pathway, waits four to six weeks for an SSRI or mood stabilizer to reach therapeutic effect. Many of those patients lose jobs, marriages, and sometimes their lives during that wait. With a similar metabolic and anti-inflammatory protocol, we see meaningful response in days, if not hours. Not because we are bypassing the diagnosis. Because we are treating the substrate the diagnosis ultimately rides on.

Insulin Resistance and the Blood-Brain Barrier

A 2021 paper in Frontiers in Psychiatry by Calkin and colleagues laid out a framework that has been building force in the bipolar literature. Insulin resistance is present in more than half of bipolar patients, and the severity of blood-brain barrier leakage tracks with both the severity of insulin resistance and the severity of bipolar illness itself. This is not a coincidence of two co-occurring metabolic problems. Insulin resistance drives a chronic inflammatory state that physically degrades the tight junctions of the blood-brain barrier, letting peripheral cytokines, immune cells, and metabolic byproducts cross into brain tissue where they do not belong. Over time, this widening leak appears to drive neuroprogression, the well-documented worsening of bipolar episodes over the disease course, with shorter euthymic intervals, more treatment resistance, and accumulating cognitive deficits.

That framework reorganizes how we think about long-term bipolar care. If the barrier is leaking and the driver is metabolic, then treating the metabolic substrate is not an adjunct. It is upstream therapy. A 2025 Lancet Psychiatry analysis from the MOBILITY trial established metformin as a new standard of care for managing the metabolic damage caused by second-generation antipsychotics in younger bipolar patients, and Mendelian randomization data now suggest metformin exposure may itself reduce bipolar risk. A growing 2025 literature in Molecular Psychiatry and elsewhere is making the same case for GLP-1 receptor agonists, which cross into the brain, reduce neuroinflammation, improve central insulin sensitivity, and modulate the HPA axis. We are starting to incorporate both where indicated, alongside oral support through Action Potential Supplements' Berberine, which improves insulin sensitivity at lower cost and with a strong evidence base of its own.

The Transition Off the IV

Acute IV therapy is an amazing rescue, but it's not practical as an ongoing solution. As the crisis settles, patients transition to oral neuronutrients drawn from the same biology, with formulations available through Action Potential Supplements — including targeted mitochondrial support through NAD+ and NMN — and in many cases they remain on a conventional mood-stabilizing medication. The distinction is that the conventional medication is often working at a lower dose, with a cleaner side effect profile, because the underlying neuroinflammation and energy failure are no longer dragging the brain backward. Mood stabilizers do their job better when the brain they are stabilizing is metabolically optimized. The recent evidence that insulin resistance breaks down the blood-brain barrier is of particular interest to us, and we are starting to incorporate strategies such as metformin and even GLP-1 receptor activating medications into our treatment paradigm. It's an exciting time to see what we have intuited for years finally being proved in the literature.

This is the model the Intensive Brain Health Program was built around. Diagnostics first, including the relevant inflammatory and metabolic biomarkers. A protocol matched to what is actually broken. A transition plan that does not abandon conventional psychiatry but builds underneath it.

Where Conventional Care Got It Right, and Where It Did Not

It would be wrong to read any of this as anti-psychiatry. Lithium remains a remarkable drug. Anticonvulsant mood stabilizers save lives. Antipsychotics have a place. What conventional inpatient care often misses is that the neurochemical conversation it has with the brain is happening on top of a substrate that has been quietly burning for years. Treat the substrate, and the conversation gets easier. Skip it, and you spend the rest of the patient's life trying to drug around a metabolic and inflammatory problem that was never going to respond to dopamine antagonism alone.

The right framing is integrative. Conventional psychiatry handles the receptor-level conversation. Neurometabolic and neuroinflammatory medicine handles the substrate. Together, they give the prefrontal cortex what it needs to come back online and resume its job as the brain's regulatory governor. That is the patient our model is designed for, and it is the patient we are seeing recover faster, with fewer hospital admissions and cleaner medication profiles, than the system on its own was ever going to deliver.