Insights
Week of 11 October 2026. Six published items. Each one is a pointer. Read the source.
A MONTH OF KETO, THEN AN OPEN FOLLOW-UP
What it is: a one-month trial, then a follow-up with no comparison group.
What it isn't: proof that keto treats schizophrenia or bipolar disorder.
Summary: fifty-eight people with schizophrenia-spectrum or bipolar I disorder were split for one month. One group ate a ketogenic diet. The other ate as usual. The keto group lost more weight. Their HbA1c and insulin resistance were also lower, though only at the edge of statistical significance. Twenty-five people then stayed on keto for four more months. Symptoms and thinking scores improved, and higher ketones went with better depression scores. Weight loss did not explain that change.
Link: Abram et al., Schizophrenia Bulletin, 2026
Takeaway: the weight change is the solid part, and the blood markers lean the same way. The mind change came after the comparison ended, so it is a signal worth following, not a result.
THE CAUTIOUS READ OF THE SAME IDEA
What it is: a review of the human studies.
What it isn't: a new trial.
Summary: most human work on keto for depression, bipolar disorder, and schizophrenia has no control group. The one trial in treatment-resistant depression, 88 people, did not clearly beat a control diet. Ketone levels did not track the scores.
Link: Oduguwa, Smith and Rakshasa-Loots, review, Current Treatment Options in Psychiatry, 2026
Takeaway: read this next to the Abram paper so one hopeful result does not stand alone.
A CLINICIAN GUIDE, NOT A NEW EXPERIMENT
What it is: an expert consensus from the Society of Metabolic Health Practitioners. That is published opinion.
What it isn't: a graded review or a new trial. The food list includes vegetables.
Summary: a panel said a whole-food lower-carb way of eating can be enough, safe, and useful for metabolic problems. They also said the early adjustment is usually mild and short. They did not grade the evidence.
Link: Society of Metabolic Health Practitioners consensus, Journal of Metabolic Health, 2026
Takeaway: this shows how some clinicians frame the work. Agreement is not proof.
KETO MEALS AGAINST A DIABETES-GUIDELINE MENU
What it is: a short preprint, not yet reviewed. A branded meal service supplied the keto arm, and two authors are tied to that company.
What it isn't: a finished paper, or a reason to buy a product.
Summary: seventeen people ate delivered meals for two five-day stretches. One was ketogenic, under 50 grams of carbohydrate a day. The other followed a menu like the American Diabetes Association pattern, about 190 grams a day. The authors report that the keto days lowered average glucose by 20 mg/dL from where people started, and the guideline menu showed no clear change. These figures have not been through peer review yet.
Link: Soto-Mota et al., preprint, Research Square, 2026
Takeaway: a head-to-head is the right kind of test. Wait for review before you lean on it.
WHAT PEOPLE WITH PARKINSON'S SAID
What it is: 16 interviews before and after 12 weeks.
What it isn't: a trial, and it has no measured movement scores.
Summary: sixteen adults with Parkinson's said they felt clearer, moved through the day better, and joined in more after a low-carb, higher-fat ketogenic program. That is their report.
Link: White, White and Tidman, interview study, Journal of Metabolic Health, 2026
Takeaway: lived experience is worth hearing. It is not a disease result.
ONE PERSON'S CARNIVORE CASE
What it is: one retrospective case report, published in a journal.
What it isn't: a pattern. Other life changes were in the mix.
Summary: a 48-year-old woman says lifelong anxiety, panic, binge eating, hot flushes, and reflux eased on a carnivore ketogenic diet. Her lipids moved toward a lean-mass-hyper-responder pattern.
Link: Newiss, case report, Journal of Metabolic Health, 2026
Takeaway: one life can point at a question. It cannot answer it.
Educational only. Not medical advice. Do some thoughtful research. Your health, your responsibility.