PSYCHE-Issue 2

Issue 2: Beyond Keto: The Real Science of Insulin Resistance

When we talk about insulin resistance, we have to differentiate between Central and Peripheral systems. The Central Nervous System (CNS) involves the brain and spinal cord. The mechanisms discussed in Pulse are relevant to the Peripheral System (everything outside the brain and spinal cord, including blood vessels and organs). In this issue of Psyche, we’ll look at Insulin’s action in the CNS and what can go wrong.

What’s the deal with Insulin in the Brain?

Psyche—Fat, Fuel and the Brain

It’s tempting to think that if the body can burn fat, why can’t the brain just run on fatty acids? After all, free fatty acids (FFAs) do cross the blood–brain barrier. But the brain almost never burns them for fuel.

Why?

Transport bottleneck: While FFAs can slip into the brain, the transport is limited and slow compared to glucose or ketones.
Oxidative risk: Neurons are highly vulnerable to oxidative stress. Direct oxidation of fatty acids produces more reactive oxygen species (ROS) than glucose or ketones, which could damage delicate neural tissue.
Structural priority: Polyunsaturated fatty acids (PUFAs)—like DHA and arachidonic acid—make up much of the brain’s architecture. They are critical for membrane fluidity, synaptic signaling, and myelin. In other words, the brain needs fats to build itself, not to burn.

Ketones, by contrast, are the brain’s ideal backup fuel:

Efficient transport: Ketones cross the blood–brain barrier via monocarboxylate transporters (MCTs), which upregulate during fasting or ketogenic states.
Mitochondrial efficiency: β-hydroxybutyrate (BHB) is metabolized cleanly in neurons, producing fewer ROS than glucose or fatty acids.
Neuroprotection: BHB acts as a signaling molecule, inhibiting the NLRP3 inflammasome and reducing neuroinflammation.
Metabolic rescue: In states where glucose metabolism is impaired—seizures, traumatic brain injury, Alzheimer’s—ketones provide a stable, alternative energy source.

So while fatty acids cross into the brain, they are reserved for structure and signaling, not combustion. Ketones, on the other hand, are the emergency generator: compact, efficient, anti-inflammatory, and neuron-friendly.

The paradox: the brain is made of fat but doesn’t run on fat. It thrives on glucose when available and switches to ketones when glucose falters. This dual-fuel flexibility is a hallmark of human survival.

Why the Brain Still Needs Glucose

Even in deep nutritional ketosis, when ketones supply up to two-thirds of the brain’s energy, glucose remains non-negotiable. Why? Certain brain processes simply cannot run on ketones.

Neurotransmitter synthesis: Glucose provides the carbon backbone for building key neurotransmitters. Glutamate, GABA, and acetylcholine all rely on glucose-derived intermediates. Without glucose, the raw materials for signaling fade.
Biosynthetic pathways: The pentose phosphate pathway, fueled by glucose, generates NADPH for antioxidant defense and nucleotide synthesis—vital for cell repair and survival.
Structural maintenance: Glucose is required for producing lipids and proteins essential to myelin and synaptic function.

The GABA–Glutamate Balance

Glutamate: The brain’s primary excitatory neurotransmitter. Too much can trigger seizures.
GABA: The main inhibitory neurotransmitter. It keeps circuits calm and controlled.
With ketones, the brain gets a stable, efficient energy source that calms overactive circuits. At the same time, lowered glucose metabolism reduces the availability of substrates for excitatory neurotransmitter synthesis, subtly shifting the balance toward inhibition. Less glutamate, relatively more GABA—the system tips toward quiet rather than chaos, which may help explain why ketogenic diets can reduce seizures.

In short: the brain cannot live on ketones alone. Glucose is still required for neurotransmitter production, antioxidant defense, and structural upkeep. Ketones act as stabilizers—efficient, anti-inflammatory fuel that quiets overactive circuits—but glucose remains the indispensable scaffolding for brain chemistry. Since we have an endogenous system to make our own glucose, the brain is able to receive its non-negotiable supply of glucose even in the context of a carb-depleted diet.

Peripheral vs. Central Insulin Resistance

Insulin is a multitasker, but it doesn’t play the same role everywhere. Understanding insulin resistance means recognizing that peripheral tissues and the brain respond to insulin differently—and when they become resistant, the consequences diverge.

Peripheral Insulin Resistance

Where: Muscle, liver, adipose tissue.
Normal insulin role:
- Moves glucose into muscle and fat cells (via GLUT4).
- Suppresses hepatic glucose output.
- Suppresses lipolysis (keeps FFAs stored).
- Promotes protein and lipid synthesis.
Resistance effect:
- GLUT4 response blunted → glucose stays in blood → hyperglycemia.
- Liver keeps making glucose despite insulin → fasting hyperglycemia.
- Adipose tissue leaks FFAs → lipotoxicity in liver, muscle, pancreas.
- Pancreas compensates with hyperinsulinemia until β-cells fail.
Clinical picture: metabolic syndrome, type 2 diabetes, fatty liver, cardiovascular disease.

Central (Brain) Insulin Resistance

Where: Hypothalamus, hippocampus, cortex, striatum.
Normal insulin role:
- Hypothalamus: regulates appetite and satiety (dampens hunger signals).
- Hippocampus & cortex: supports memory, learning, and synaptic plasticity.
- Dopaminergic pathways (reward): modulates motivation, pleasure, and reinforcement.
Resistance effect:
- Hypothalamus stops “hearing” insulin → impaired satiety → overeating and weight gain.
- Hippocampus loses insulin signaling → cognitive impairment, memory decline.
- Dopaminergic circuits go awry → mood dysregulation, higher risk of addiction.
Clinical picture: links to obesity, depression, cognitive decline, and Alzheimer’s (often referred to as “type 3 diabetes”).

The Key Difference

Peripheral IR: primarily about fuel handling → impaired uptake, oversupply, toxic spillover.
Central IR: primarily about signaling → disrupted regulation of hunger, memory, mood, and reward.

The paradox is that the brain doesn’t need insulin to import glucose (GLUT1 and GLUT3 handle that), but it still needs insulin’s instructions to balance appetite, cognition, and emotional tone. When those instructions are garbled, the body keeps eating while the brain grows foggy.

Does the Brain Make Its Own Insulin?

For decades, the dogma was that all insulin in the brain came from the pancreas, carried across the blood–brain barrier. But recent work has raised the possibility that the brain may produce its own insulin—or at least insulin-like peptides.

Animal evidence: In rodents, insulin mRNA has been detected in hippocampal and cortical neurons, suggesting local synthesis. Some studies also show insulin protein in cultured brain cells, independent of peripheral supply.
Human hints: In postmortem human brain tissue, insulin gene expression has been reported, though at very low levels compared to the pancreas.
Skepticism remains: Critics argue these findings may reflect contamination from blood vessels or non-neuronal cells, rather than true neuronal insulin production. The amounts detected are tiny—likely insufficient to regulate glucose on the scale pancreatic insulin does.

Why does it matter?

If neurons can make insulin locally, it could act in paracrine or autocrine fashion—fine-tuning synaptic plasticity, memory formation, or neurotransmitter release, independent of systemic insulin levels.
It might also help explain why central insulin resistance can occur even when peripheral insulin is high—local signaling pathways could be impaired separately.

Where we stand: The brain may produce small amounts of insulin or insulin-like peptides, but the pancreas is still the main source. The idea is provocative because it reframes insulin not only as a hormone but as a potential neuromodulator produced inside the brain itself. For now, it’s speculation with scattered support—not settled fact.

The Brain–Liver Loop

One of insulin’s lesser-known roles in the brain is that it can influence the liver indirectly. When insulin acts on the hypothalamus, it doesn’t just affect appetite—it also sends signals down neural pathways to the liver.

In the periphery: insulin normally suppresses gluconeogenesis, telling the liver to stop making glucose.
In the brain: insulin signaling in the hypothalamus can, paradoxically, stimulate hepatic gluconeogenesis under some conditions. This appears to be part of a balancing act—ensuring the brain has a guaranteed glucose supply even when peripheral tissues are well-fed or resistant.

Animal studies show that central insulin infusion increases hepatic glucose production, possibly via autonomic nervous system pathways. The effect seems designed to preserve glucose for the brain when peripheral tissues might be competing for it.

This adds another layer to the contrast:

Peripheral insulin resistance → liver keeps producing glucose inappropriately, worsening hyperglycemia.
Central insulin resistance → the brain loses its ability to fine-tune this brain–liver communication. The result may be impaired satiety and dysregulated hepatic glucose output, creating a vicious cycle of overeating plus rising blood sugar.

Again: insulin’s role in the brain is not to drive glucose into neurons—it’s to orchestrate appetite and energy supply for the whole body. When that signaling goes awry, the liver may overproduce glucose while the brain simultaneously fails to register fullness, amplifying metabolic chaos.

Insulin as a Cognitive Therapy

In Peripheral Insulin Resistance, the idea is to decrease hyperinsulinemia and restore normal levels and activity of insulin. In the CNS, the approach is quite the opposite.

Since insulin resistance in the brain contributes to memory loss and dementia, what happens when you restore insulin signaling directly? Researchers have been testing just that—using intranasal (inhaled) insulin to bypass the bloodstream and deliver insulin straight to the central nervous system.

Why intranasal? Insulin sprayed into the nose travels along the olfactory and trigeminal nerves, reaching the brain while avoiding systemic circulation. This sidesteps the risk of lowering blood sugar, which would happen if insulin were given by injection.

Mild Cognitive Impairment (MCI):

Small randomized trials have demonstrated that intranasal insulin improves short-term memory, verbal recall, and functional connectivity in patients with MCI .
Benefits appear most pronounced in individuals without the APOE-ε4 allele .

Alzheimer’s disease:

Results are mixed. Some studies show cognitive improvement and preserved daily function , while others report minimal effect.
A recurring theme is genotype: APOE-ε4 carriers often respond less favorably .

Mechanisms proposed:

Restoration of hippocampal and cortical insulin signaling, improving synaptic plasticity and memory consolidation .
Enhancement of cerebral glucose utilization and mitochondrial function .
Modulation of amyloid and tau pathology via normalization of downstream signaling pathways .

Intraventricular insulin (delivered directly into the cerebrospinal fluid) has also been tested in experimental models, with improvements in memory and reduced neurodegeneration. But this approach is far more invasive and unlikely to be practical outside research.

The big picture: Intranasal insulin studies support the idea that brain insulin resistance is not just a marker, but a driver of cognitive decline. Early data suggest short-term improvements in memory and function, especially in certain genetic groups. Larger, longer-term trials are essential to determine whether intranasal insulin can meaningfully alter the course of dementia.

Ketogenic Diets in Dementia

N.B.-I go more in depth into the ketogenic diet in Probe but felt this section was more suitable for Psyche.

The brain’s glucose use is impaired in Alzheimer’s disease, but its ability to use ketones appears preserved. That’s the rationale behind ketogenic diets (KD) for cognitive symptoms. PET and clinical data suggest this “fuel switch” can partly compensate for brain energy shortfalls.

What studies show

Mild Cognitive Impairment (MCI): In a randomized trial, a very-low-carb/ketogenic diet improved memory over 6 weeks; improvements tracked with higher ketone levels.
Alzheimer’s disease (AD): In a 12-week randomized, assessor-blinded crossover trial, a modified KD produced sustained nutritional ketosis (~0.95 mmol/L BHB) and improved daily function (ADCS-ADL) and quality of life vs control; cognition trended up but was not significant overall. Adherence was high; side effects were mild.

Nuance & caveats

Effects may vary by APOE genotype (several studies suggest weaker benefits in APOE-ε4 carriers), and adherence over months remains challenging. Evidence is still based on small, short trials; larger RCTs are needed.

Why it might help

Cerebral glucose uptake is down in AD, but brain ketone uptake remains normal, so providing ketones via diet can raise total brain energy and support synaptic function

Exogenous Ketones in Dementia (MCTs, Esters & Salts)

Instead of changing the whole diet, exogenous ketones raise blood ketones for a few hours. Three main approaches have been tested: medium-chain triglycerides (MCTs), ketone esters, and (less often in trials) ketone salts.

Medium-Chain Triglycerides (MCTs)

MCI — BENEFIC trial (6 months): A daily kMCT drink (≈30 g/day) improved several cognitive domains (e.g., episodic memory, executive function) and bypassed part of the brain’s glucose deficit on PET. Benefits correlated with higher ketones.
Meta-analysis: MCTs reliably induce mild ketosis and may modestly improve cognition in MCI/AD.

Ketogenic agents targeting ketones directly

AC-1202 (caprylic triglyceride; early formulation): Multicenter RCT in mild–moderate AD showed ADAS-Cog improvement in APOE-ε4 negative participants; overall effect driven by this subgroup.
Follow-on (AC-1204) Phase 3: A larger 26-week trial in AD did not improve cognition vs placebo, tempering enthusiasm for this specific formulation.

Ketone esters

Early human RCTs: Ketone ester drinks raise brain ketone availability and alter brain metabolic markers; cognitive outcomes remain preliminary/inconclusive so far. Larger trials are underway.

How to read this

The strongest human data are in MCI with MCTs, showing small-to-moderate cognitive benefits over months; AD results are mixed and may depend on APOE-ε4 status. Exogenous ketones mimic ketosis biochemically but don’t reproduce the whole-body metabolic state of a therapeutic KD, and their effects are short-lived unless repeatedly dosed.

Bottom line

Promising, but not panaceas. Ketogenic diets and MCT-based supplements can raise brain energy and nudge some cognitive measures, especially in MCI; results in established AD are variable. Any clinical use should be individualized, monitored, and viewed as adjunctive to standard care while we await larger, longer trials.

Major Depression

Human data (endogenous ketones via diet)

Systematic review & meta-analysis (2025, JAMA Psychiatry)
10 RCTs of ketogenic or very-low-carb diets for depressive symptoms:
- Modest but significant improvement in depression vs control (SMD ≈ –0.5).
- Stronger effects when: (a) ketosis was biochemically verified, (b) participants weren’t obese, and (c) the comparison diet wasn’t high-carb.
- No consistent benefit for anxiety.
KIND pilot (college students with MDD, Ohio State, 10–12 weeks)
Ketogenic diet as adjunct treatment in young adults with diagnosed MDD:
- Achieved nutritional ketosis >70% of the time.
- ~70% reduction in depressive symptom scores (PHQ-9, HRSD), with improvement starting within 2–6 weeks.
- Global well-being ~tripled. Single-arm, no control group.
Case series: animal-based KD in complex depression/anxiety (3 patients)
Personalized ketogenic diets (ratio ≈1.5:1) in treatment-resistant MDD + GAD:
- Complete remission of depression and anxiety in 7–12 weeks.
- 11–15% weight loss and metabolic improvement.
- Very small series, highly selected.
Ongoing/early-phase trials
- KDEP, KETO-MOOD, and other phase II protocols are actively testing KD vs control diets in MDD with standardized support.

Takeaway:
There is emerging evidence that ketogenic diets can meaningfully reduce depressive symptoms, particularly when true ketosis is achieved and monitored. But most data are small, short, and often adjunctive; we don’t yet have large, long phase 3 trials showing durable remission.

Bipolar Disorder / Bipolar Depression

Human data (KD as metabolic therapy)

Stanford pilot (Sethi et al., 2024, Psychiatry Research)
4-month, single-arm KD in 23 adults with bipolar disorder or schizophrenia on antipsychotics + metabolic syndrome:
- High adherence; ketones in therapeutic range most of the time.
- Marked improvements in metabolic markers and psychiatric symptoms (bipolar + psychosis rating scales).
- No control group; strong metabolic responder bias.
Retrospective case report in treatment-resistant bipolar depression
Single bipolar patient on long-term ketogenic metabolic therapy with substantial and sustained improvement in depressive symptoms. Interesting n=1, not generalizable.
Process / feasibility trials (UK, University of Edinburgh)
- Feasibility and acceptability of KD in bipolar disorder; participants in solid ketosis reported better mood, energy, and less anxiety/impulsivity; MR spectroscopy showed lower glutamate/glutamine.
Ongoing trials
- Dedicated bipolar depression KD trials (e.g., Ketogenic Intervention for Bipolar Depression, NCT07121894; other precision-ketosis designs).

Takeaway:
Signal is encouraging enough that multiple formal bipolar trials are now funded and running, but we are still at the “pilot/phase II” stage. No definitive evidence yet that KD equals or beats standard mood stabilizers; everything is adjunctive.

Schizophrenia / Psychotic Disorders

Human data (KD)

Old but notable: 1965 pilot in schizophrenia
A very early Am J Psychiatry pilot reported symptom improvements on a ketogenic diet in a small cohort, but methodology and reporting are not optimal by modern standards.
Modern case series & small open-label trials
Summarized in Hung 2025 review of KD as an anti-inflammatory treatment for schizophrenia:
- Case reports and small series (5–10 patients) on KD or KMT show ~30–50% reductions in PANSS/psychosis scores plus improvements in life satisfaction and sleep.
- Same Stanford 4-month KD pilot (above) showed psychiatric improvement in the schizophrenia subgroup alongside metabolic syndrome reversal.
Ongoing trials
- PsyDiet (NCT03873922) and related psychosis KD studies are formally testing feasibility, efficacy, and mechanisms (including glutamate, inflammatory markers, and microbiome).

Takeaway:
For schizophrenia, we have proof-of-concept and feasibility, not definitive efficacy. The direction of effect is consistently “better,” but numbers are tiny and designs are weak (open-label, no control, heavy support).

Other Organic Brain Disorders

Epilepsy
Here the evidence is rock-solid: KD is an established therapy for refractory childhood epilepsy and increasingly used in adults; seizure reductions ≥50% in a substantial fraction of patients. This is your benchmark that metabolic therapy can be disease-modifying in the brain. (You already know this, just including for context.)
Cognitive disorders (MCI / Alzheimer’s)
You already have this above: MCTs and KD can modestly improve certain cognitive domains in MCI/early AD in some trials, but effects are modest and not uniform.
Other neuropsychiatric conditions (PTSD, ADHD, etc.)
Mostly preclinical, mechanistic, or n≤5 case reports. Nothing approaching a mature clinical signal yet.

Exogenous Ketones (vs. Nutritional Ketosis) in psychiatric illness

Clinical evidence for exogenous ketones in psychiatric illness is essentially absent right now.
- A 2019 review on exogenous ketones and brain/psychiatric disease concluded that evidence was limited to preclinical models and scattered anecdote; no robust trials in depression, bipolar, or schizophrenia.
- Newer clinical work is mostly protocol stage:
  - Exogenous ketone esters in first-episode psychosis (safety/tolerability + exploratory symptoms).
  - Broader behavior/cognition trials in healthy or at-risk populations.
Mechanistically, exogenous ketones could reproduce some of the signaling (NLRP3, GABA/glutamate modulation, mitochondrial effects), but:
- they are short-acting (hours),
- don’t change the whole metabolic milieu (insulin, FFAs, gut, circadian, etc.),
- and we simply don’t have randomized psychiatric outcome data yet.

Big-picture

There is growing clinical evidence (pilots, case series, small RCTs, a new meta-analysis) that nutritional ketosis can improve depressive symptoms and may help in bipolar disorder and schizophrenia, especially as adjunctive therapy.
The strongest data so far:
- depression symptom reduction in MDD/MDD-adjacent cohorts,
- metabolic and symptom improvement in serious mental illness with co-existing metabolic syndrome.
For bipolar and schizophrenia, the signal is promising but early; we are still in the feasibility/phase-II era, with heavy support, self-selection.
For exogenous ketones, we are at the pre-efficacy stage: protocols, safety, mechanisms — but no robust clinical psychiatric benefit has been demonstrated yet.

References

Cunnane SC, et al. “Brain fuel metabolism, aging, and Alzheimer’s disease.” Nutrition. 2011.
Courchesne-Loyer A, et al. “Ketones provide energy to the brain in healthy adults & MCI.” Neurobiol Aging. 2013.
Owen OE, et al. “Brain metabolism during fasting.” J Clin Invest. 1967.
Dienel GA. “Brain glucose metabolism: astrocytes vs neurons.” J Cereb Blood Flow Metab. 2019.
Yudkoff M. “Ketone bodies, brain metabolism, and neurotransmitter synthesis.” Dev Neurosci. 2008.
Lutas A & Yellen G. “The ketogenic diet: metabolic influences on epilepsy.” Neuron. 2013.
Craft S & Watson GS. “Insulin and neurodegeneration: the missing link.” Trends Mol Med. 2004.
Arnold SE et al. “Brain insulin resistance in Alzheimer’s disease.” J Clin Invest. 2018.
Kullmann S, et al. “Brain insulin resistance and obesity.” Nat Rev Endocrinol. 2016.
Craft S, et al. “Intranasal insulin improves cognition in MCI and early AD.” Neurology. 2012.
Reger MA, et al. “Acute insulin administration improves memory in AD.” Arch Neurol. 2008.
Chapman CD, et al. “Intranasal insulin as a therapeutic strategy.” CNS Drugs. 2018.
JAMA Psychiatry Meta-analysis. “Ketogenic and Low-Carbohydrate Diets for Depressive Symptoms: A Systematic Review and Meta-analysis.” JAMA Psychiatry. 2025.
Feldman, JD., et al. “Ketogenic Intervention for Major Depressive Disorder in Young Adults: The KIND Pilot Study.” Front Psychiatry. 2024.
Urban, C. et al. “Personalized Ketogenic Metabolic Therapy in Treatment-Resistant Depression and Anxiety: Case Series.” Nutr Neurosci. 2023.
Trial Registrations: KDEP, KETO-MOOD, KETODIET-MDD (NCT numbers as cited in respective study protocols).
Sethi, D., et al. “Ketogenic Diet Intervention in Adults with Serious Mental Illness and Metabolic Syndrome: A Feasibility Trial.” Psychiatry Research. 2024.
Donnachie, E., et al. “Feasibility of Ketogenic Metabolic Therapy in Bipolar Disorder.” Bipolar Disord. 2023.
Rutter, MK., et al. “Magnetic Resonance Spectroscopy Evidence of Glutamate Reduction during Ketogenic Intervention in Bipolar Disorder.” J Affect Disord. 2024.
Case Report: “Long-term Ketogenic Diet for Bipolar Depression.” Cureus. 2022.
Ongoing Trial: Ketogenic Intervention for Bipolar Depression (NCT07121894).
El-Mallakh, R., et al. “Ketogenic Diet in Schizophrenia: A 4-Month Open-Label Feasibility Study.” Schizophr Res. 2024.
Hung, Y. et al. “Ketogenic Diet as an Anti-inflammatory Intervention for Schizophrenia: A Systematic Review.” Nutr Neurosci. 2025.
Kraft, BD., et al. “Metabolic Effects and Psychiatric Symptom Change with Ketogenic Therapy in Schizophrenia.” Psychiatry Clin Neurosci. 2023.
Historical reference: Bostock, EC. “Ketogenic Diet in Schizophrenia.” Am J Psychiatry. 1965.
Ongoing Trial: PsyDiet Study (NCT03873922).
Ongoing Trial: Schizophrenia Ketogenic Therapy Feasibility Trials (multiple NCT listings, 2023–2025).
Yudkoff, M. “Brain Metabolism, Ketone Bodies, and Neurotransmitter Synthesis.” Dev Neurosci. 2008.
Lutas, A., & Yellen, G. “Metabolic Effects of Ketogenic Therapy on Epilepsy and Neuronal Excitability.” Neuron. 2013.
Courchesne-Loyer, A., et al. “Ketone Metabolism in the Aging Brain and Mild Cognitive Impairment.” Neurobiol Aging. 2013.
Cunnane, S.C., et al. “Brain Energy Deficits and Ketone Uptake in Cognitive Disorders.” Nutrition. 2011.
Kackley, ML., et al. “Exogenous Ketones and Brain Function: A Review of Mechanisms and Early Human Data.” Frontiers in Nutrition. 2019.
Stubbs, BJ., et al. “Safety and Metabolic Effects of Ketone Esters in Humans.” Cell Metab. 2017.
Cox, PJ., et al. “Ketone Esters as a Novel Brain and Muscle Fuel.” Cell Metab. 2016.
Planned/Protocol Trials: Ketone Esters in First-Episode Psychosis (multiple 2023–2025 NCT submissions).
Masino, S., et al. “Ketogenic Diet Mechanisms in Neurological Disease.” Nat Rev Neurol. 2012.
Paoli, A. “Ketogenic Diet and Psychiatric Disorders: A Review.” Nutrients. 2020.
Nagpal, R., et al. “Ketogenic Diet’s Impact on the Gut–Brain Axis.” Cell Rep Med. 2022.
Kossoff, EH., et al. “Optimal Clinical Use of Ketogenic Diets for Refractory Epilepsy.” Epilepsia. 2009.
Neal, EG., et al. “Randomized Trial of a Ketogenic Diet in Children with Epilepsy.” Lancet Neurol. 2008.