What the Research Actually Says

If you read carnivore forums you will see strong claims in both directions: that an all-meat diet cures almost everything, or that it will inevitably damage your heart and gut. Neither claim is supported by good evidence. This article lays out what research actually exists, what type of evidence each piece is, and how much weight it can bear.

A quick guide to evidence quality

Not all studies are equal. From strongest to weakest for answering "does this diet cause this outcome?":

1. Randomized controlled trials (RCTs): people are randomly assigned to diets, so differences in outcome are more likely to be caused by the diet. Long RCTs with hard outcomes (heart attacks, deaths) are the gold standard, and are rare in nutrition.
2. Controlled non-randomized trials: useful, but people who choose a diet may differ from those who do not.
3. Observational (cohort) studies: follow large numbers of people over time; good for spotting associations, weak for proving cause.
4. Surveys and case series: describe what people report; cannot establish cause, and are prone to selection bias.
5. Mechanistic reasoning: what should happen based on physiology; useful for hypotheses, often wrong in practice.
6. Anecdote: individual stories; valuable for generating questions, not for answering them.

For the carnivore diet specifically, almost all the direct evidence sits in categories 4 to 6.

The 2021 carnivore survey (Lennerz et al.)

The largest study of people eating a carnivore diet is an online survey of 2,029 adults who had followed the diet for at least six months, published in Current Developments in Nutrition^[1]. Participants were recruited mainly through social media and carnivore communities.

What it found

• Median age was 44, about two-thirds were men, and the median time on the diet was about 14 months.
• Respondents reported very high satisfaction and improvements in overall health, energy, sleep, and a range of other measures.
• Self-reported median BMI fell from about 27 to about 24.
• Among respondents with diabetes, most reported reducing or stopping diabetes medications, along with lower HbA1c.
• Reported adverse symptoms were uncommon, with gastrointestinal complaints the most frequent.
• Among those who shared lab results, LDL cholesterol was high (median around 170 mg/dL), while HDL cholesterol was high and triglycerides low.

What it cannot tell us

The authors themselves acknowledged important limitations, and a published critique highlighted them further^[2]:

• Selection bias. People who felt worse on the diet and quit were largely not captured.
• Self-report. Weight, lab values and health status were reported by participants, not measured.
• No control group. Improvements could come from weight loss, removing ultra-processed food, or increased health focus, not from eliminating plants.
• Short duration relative to the timescale of heart disease or cancer.

Fair verdict: the survey shows that a sizable group of people can follow a carnivore diet for over a year and feel well doing it, and it generated useful hypotheses. It does not show that the diet is safe or effective for people in general.

The Bellevue experiment (1928 to 1929)

The only long-term supervised study of an all-meat diet in modern times involved just two men: the Arctic explorer Vilhjalmur Stefansson and his colleague Karsten Andersen. They ate only meat, fat and organs for a year under the supervision of a scientific committee, starting with a period of close observation at Bellevue Hospital in New York.

The main scientific report, by McClellan and Du Bois in the Journal of Biological Chemistry (1930), focused on kidney function and ketosis and found no evidence of kidney damage over the year; both men were in mild ketosis throughout^[3]. A clinical report by Lieb in JAMA described the men as remaining in good health^[4]. Neither developed scurvy.

This was careful work for its time, but it involved two healthy men for one year. It shows that an all-meat diet including fat and organs is possible without obvious short-term harm; it cannot tell us about long-term risks in the wider population. We cover the story in detail in Vilhjalmur Stefansson and the Bellevue Experiment.

What low-carbohydrate and ketogenic trials add

Because carnivore is almost always very low in carbohydrate, research on ketogenic and low-carbohydrate diets is the most relevant body of controlled evidence. It is much larger and includes RCTs.

• Type 2 diabetes: a 2021 systematic review and meta-analysis in The BMJ found that low-carbohydrate diets led to higher rates of diabetes remission at six months compared with control diets, with the benefit less clear by twelve months^[5].
• Intensive support programs: a non-randomized study of a ketogenic diet delivered with remote medical coaching reported substantial HbA1c reductions and large reductions in diabetes medication at one year^[6].
• Appetite and energy intake: a tightly controlled NIH study comparing an animal-based ketogenic diet with a plant-based low-fat diet found people spontaneously ate fewer calories on the low-fat diet, while the ketogenic diet produced lower glucose and insulin levels^[7].

These trials support some of the metabolic effects people report on carnivore. But they tested diets that included plants, so they are indirect evidence for carnivore, not proof.

What reviews of the carnivore literature conclude

A 2026 scoping review in Nutrients identified only nine human studies of the carnivore diet, and characterised the evidence as limited by small sample sizes, short durations and lack of control groups^[8]. The authors concluded the diet may offer short-term benefits but raised concerns about potential nutrient shortfalls and elevated LDL cholesterol, and did not recommend it long term on current evidence. That is a cautious, mainstream reading, and it is fair to note that the absence of evidence cuts both ways: we also lack good evidence of long-term harm specific to this diet.

Evidence map

Question	Best available evidence	Confidence
Can people follow carnivore for a year and feel well?	Survey, Bellevue experiment, case reports	Moderate that some can; unknown how many
Does very-low-carb eating improve blood glucose?	RCTs and meta-analyses (low-carb, not carnivore)	Reasonably high, short to medium term
Does it help weight loss?	RCTs of low-carb/high-protein diets	Works for many, not clearly superior long term
Does it improve autoimmune, mood or skin conditions?	Surveys and anecdote	Low
Does it raise LDL cholesterol?	Survey data, keto RCTs	Often, sometimes a lot; varies widely
Long-term heart disease, cancer, mortality?	None specific to carnivore	Unknown

What would settle the question

We need randomized trials comparing a well-formulated carnivore diet with a well-formulated alternative (for example, a Mediterranean or low-carb diet with vegetables), lasting at least a year, with measured rather than self-reported outcomes, and careful tracking of lipids, micronutrients, and gut health. Until then, anyone following this diet is effectively running an experiment on themselves, and should track their own data with their doctor.