Dr. Georgia Ede
Harvard-trained psychiatrist · nutritional/metabolic psychiatry
The genuinely credentialed and most careful figure in the low-carb/metabolic-psychiatry space — she has real peer-reviewed (consensus/review) publications and clears the bar. Her narrow, mechanistic claims (ketones are a real alternative brain fuel; plant antinutrients reduce mineral absorption) are defensible. But her headline claims that ketogenic diets meaningfully treat serious mental illness, depression and anxiety currently outrun the evidence: the best-controlled 2026 data (a JAMA Psychiatry RCT and meta-analysis) find efficacy uncertain, and the supporting pilots are uncontrolled and cluster in a single advocacy-funded network — a candor her own consensus paper shares.
Evidence vs. their claims
Across 5 testable claims, where the weight of independent evidence lands — on their side, or against them. Their own research network is excluded as evidence.
Read relative to their stance: when they dispute a claim and the evidence supports that claim, it counts as against them. Each row shows their stance, so you can see which way the evidence cuts.
Conflicts of interest
- Book (Change Your Diet, Change Your Mind) and clinician-training courses (Diagnosis Diet)
- Aligned with the Baszucki Group / Metabolic Mind advocacy-funding network that bankrolls much of the keto-for-mental-illness field, including the flagship supporting pilots
Stated factually. Conflicts do not by themselves make a claim wrong; they are context for weighing it.
Claim-by-claim (5)
Grouped by health topic. A person's claims often span several areas — here's where each one lands.
🥗 Diets 5
Can the brain run well (or better) on ketones instead of glucose?
Can a ketogenic diet meaningfully improve bipolar disorder or schizophrenia?
Does a ketogenic diet reduce depression or anxiety?
Is a meat-based diet better than a plant-based one at feeding your brain?
Do plant foods contain antinutrients that limit nutrient absorption?
Educational only, not medical advice. Verdicts grade the evidence, not the person; see the methodology.