A Cochrane review of 15 long trials and about 59,000 adults. Cutting saturated fat went with 21% fewer combined cardiovascular events, while deaths from any cause barely moved, and swapping it for polyunsaturated fat or for carbohydrate looked similar.
the research library food, supplements and pregnancy
the Ray Peat diet
A way of eating built around the thyroid. Here is what the research says about each piece of it.
Clients bring me screenshots of pro-metabolic eating rules and ask whether the science holds, and the rules are specific enough to check one at a time. So I read the trials and the cohorts myself, including the ones that cut against the idea, so you can see where each claim stands instead of taking anyone's word for it.
what the diet is
It comes from the writing of Ray Peat, a biologist who wrote newsletters and books about hormones and metabolism. People who follow it eat plenty of fruit, orange juice, honey and sugar, milk and cheese, gelatin and broth, shellfish, eggs and liver, coffee, butter and coconut oil, often with a raw carrot salad, and they avoid seed oils, fasting and very low carb eating. The thyroid, body temperature and pulse sit at the center of it.
14papers
13studies in people
6counterpoints
this page, by tier
- 10 human evidence
- 3 early human
- 1 lab + animal
- 0 background
the short version
what the research found
the headline numbers
-
In 9,423 adults in the recovered Minnesota Coronary Experiment, swapping saturated fat for corn oil lowered cholesterol 13.8% but gave no survival advantage.
-
Yet measured linoleic acid in blood and tissue went with 22% lower cardiovascular death across 30 cohorts and 68,659 adults, the strongest evidence against avoiding seed oils across the board.
-
Swapping sugar for other carbohydrates at equal calories changed body weight by 0.04 kg, while eating more sugar on top of a usual diet added about 0.75 kg.
-
Whole fruit went with slightly lower type 2 diabetes risk, while the same amount of fruit juice went with 8% higher risk, across 187,382 adults.
-
Across 29 cohorts and 938,465 adults, milk and total dairy showed no link to death from any cause, coronary heart disease or cardiovascular disease.
-
On 800-calorie diets with no carbohydrate, T3 fell 47%, while the same calories with at least 50 g of carbohydrate left T3 unchanged.
-
Body temperature is a weak thyroid readout: hypothyroidism went with a temperature just 0.013 °C lower in 35,488 patients.
the nano angle
Milk and fruit, two staples of this way of eating, both carry extracellular vesicles, and pulling vesicles out of foods like these was the work of my PhD. No study of this diet has measured food vesicles or what they do after a meal, so I hold that as an open question, not a selling point.
my honest bottom line
Parts of this way of eating hold up better than its critics allow: swapping saturated fat for linoleic acid did not lower deaths in the recovered trials, sugar swapped calorie for calorie did not change body weight, and dietary carbohydrate really does move T3. The biggest limitation is that no one has tested the whole pattern in a trial I could find, so every claim here rests on a single food or nutrient studied on its own, usually in people eating nothing like this.
evidence at a glance
14 papers
- human evidence10
- early human3
- lab + animal1
- backgroundnone yet0
Includes 6 counterpoints.
every title links to the original
the papers
Sorted from the strongest human evidence down to the lab work. Some of these support the diet and some push back hard, and both kinds stay in on purpose.
Instead of asking people what they ate, 30 cohorts measured linoleic acid in blood and tissue across 68,659 adults. Higher levels went with less cardiovascular disease, 22% lower cardiovascular death and fewer ischemic strokes, and arachidonic acid, the metabolite blamed for inflammation, was not tied to higher risk.
Across 29 cohorts and 938,465 adults, milk and total dairy showed no link to death from any cause, coronary heart disease or cardiovascular disease. Fermented dairy and cheese went with slightly lower risk until one large Swedish cohort was removed.
An umbrella review of 218 meta-analyses. Three to four cups of coffee a day went with 17% lower death from any cause and 19% lower cardiovascular death than none. Harm signals faded after accounting for smoking, except in pregnancy, where high intake went with low birth weight and pregnancy loss.
In 35,488 patients and 243,506 temperature readings, the average was 36.6 °C, and hypothyroidism went with a temperature only 0.013 °C lower. Everything measured, thyroid included, explained just 8.2% of the differences between people.
Data from a 1968 to 1973 double-blind trial, recovered decades later. Swapping saturated fat for corn oil lowered cholesterol 13.8% versus 1.0% in controls, with no survival advantage in anyone. Within the trial, each 30 mg/dL drop in cholesterol went with a 22% higher risk of death, an association, not proof of harm.
show all 14 papersshow fewer
Recovered data from a 1966 to 1973 trial in 458 men after a heart event. Men told to swap saturated fat for safflower oil died more often than controls (17.6% versus 11.8%). Added to other linoleic acid trials, the pooled result leaned toward more coronary deaths without reaching significance.
Three US cohorts, 187,382 adults followed up to 24 years. Per three servings a week, whole fruit went with slightly lower type 2 diabetes risk (blueberries 26% lower), while fruit juice went with 8% higher risk. Oranges were neutral.
Across 30 trials and 38 cohorts, adults who cut sugar lost about 0.80 kg and those who ate more gained about 0.75 kg. Swapped for other carbohydrates at equal calories, sugar changed nothing, and the highest sugary drink intakes carried 1.55 times the odds of overweight or obesity.
Across 31 controlled feeding trials at equal calories, fructose in place of other carbohydrates did not change body weight. In 10 trials that added 104 to 250 g of fructose a day on top of usual eating, weight rose 0.53 kg, which the authors put down to the extra calories.
Eight men drank 15 g of vitamin C-enriched gelatin an hour before six minutes of jumping rope. A blood marker of collagen building doubled compared with placebo, and their serum helped lab-grown ligaments gain collagen and strength.
21 adults who had lost 10 to 15% of their weight ate three diets for four weeks each. Resting burn fell least on very low carb and most on a high glycemic load low-fat diet, but very low carb also raised 24-hour urinary cortisol, so no diet won on every measure.
Fasting for 7 to 18 days cut T3 by 53% and raised reverse T3 by 58%. On 800-calorie diets with no carbohydrate, T3 still fell 47%, while at least 50 g of carbohydrate at the same calories kept T3 steady, the closest direct human test of the carbohydrate and T3 idea.
Mice fed an 8% glycine diet lived 4 to 6% longer, males and females alike, at all three test sites. Glycine is the amino acid gelatin is richest in, which is why this study gets quoted. Mice, not people.
keep reading
more from the library
food, supplements and pregnancy eating for metabolism Meal timing, protein, fiber, a walk after dinner, and the trials that found less than the headlines. food, supplements and pregnancy therapeutic diets Mediterranean, DASH, ketogenic and low-FODMAP diets as tested in real trials, retraction included. the whole person precision wellness Why two people eat the same meal and get different results, and what personalized advice has delivered so far.
This page summarizes published research for education. It is not medical advice, and nothing here describes what any nanonourish™ product does. Talk with your care provider before changing your diet or starting a supplement, especially if you are pregnant, nursing or taking medication. Summaries by Dr. Kaylan Jackson, PhD, a chemist, not a physician. Citations checked against PubMed and Crossref, September 2026.