Paleo, keto and carnivore: what the evidence says

By the DieticiansWorldwide.com editorial team · 17 October 2026 · Public health & global nutrition

Three diets, one shared idea

Paleo, keto and carnivore are often lumped together because they share a starting point: the belief that modern carbohydrate-rich food, especially grains and sugar, is behind much of today's obesity and metabolic disease. Each takes that idea further than the last.

• Paleo asks you to eat like a pre-agricultural hunter-gatherer. A 2015 systematic review defined it as vegetables, fruit, nuts, fish, meat and eggs, excluding "dairy, grain-based foods, legumes, extra sugar, and nutritional products of industry" (Manheimer and colleagues, AJCN 2015).
• Keto (the ketogenic diet) restricts carbohydrate so severely that the body switches to burning fat and producing ketones. Research definitions usually set the limit at no more than 50 g of carbohydrate a day (Bueno and colleagues, British Journal of Nutrition 2013).
• Carnivore removes plant foods almost entirely. Its followers eat meat, fish, eggs and sometimes dairy.

What the trial evidence says about paleo

The paleo evidence base is small. The 2015 meta-analysis found only four randomised trials, with 159 participants in total, all in people with at least one feature of metabolic syndrome (Manheimer and colleagues, 2015). Over six months or less, paleo did better than guideline-based control diets: waist circumference fell by an extra 2.38 cm and triglycerides by an extra 0.40 mmol/L, with smaller, less certain differences in blood pressure and blood sugar. The authors rated the evidence moderate and concluded that more research was "warranted".

The longest trial shows what happens over time. In Sweden, 70 obese post-menopausal women followed either a paleo-type diet or the Nordic Nutrition Recommendations for two years (Mellberg and colleagues, European Journal of Clinical Nutrition 2014). The paleo group lost more fat at six months (6.5 kg against 2.6 kg), but by 24 months the difference was no longer statistically significant (4.6 kg against 2.9 kg). Triglycerides stayed lower in the paleo group, and the authors noted that adherence to the planned protein intake was poor.

The pattern is common in diet research: a strong early lead that narrows as people drift back towards their usual eating. Paleo does encourage plenty of vegetables and fruit and cuts ultra-processed food, which most dietitians would support. The weaker parts are the blanket exclusions of whole grains, pulses and dairy, foods with good evidence behind them in national guidelines.

What the trial evidence says about keto

Keto has far more research than paleo, and the headline finding is modest.

• Weight over a year. A meta-analysis of 13 randomised trials lasting 12 months or more found that very-low-carbohydrate ketogenic diets produced 0.91 kg more weight loss than low-fat diets, with lower triglycerides and diastolic blood pressure and higher HDL cholesterol, but also higher LDL cholesterol (Bueno and colleagues, 2013).
• Low-carb versus low-fat, head to head. In the DIETFITS trial, 609 adults were randomised to a healthy low-fat or healthy low-carbohydrate diet for 12 months. Weight loss was 5.3 kg and 6.0 kg respectively, a difference that was not statistically significant. Neither genotype nor insulin secretion predicted who would do better on which diet (Gardner and colleagues, JAMA 2018).
• Blood sugar. The Keto-Med trial gave 40 adults with prediabetes or type 2 diabetes 12 weeks each of a well-formulated ketogenic diet and a Mediterranean-style diet, both low in added sugar and refined grains. HbA1c improved on both and did not differ between them. Keto lowered triglycerides more (16% against 5%), but LDL cholesterol rose by 10% on keto while falling by 5% on the Mediterranean diet, and fibre and several nutrients were lower on keto. Follow-up data suggested the Mediterranean diet was "more sustainable" (Gardner and colleagues, AJCN 2022).

Read together, these trials suggest that much of keto's benefit comes from what it shares with other good diets: less sugar, fewer refined grains, more vegetables and fewer total calories. Cutting carbohydrate to ketogenic levels adds little extra weight loss on average and brings a cholesterol trade-off. Our guide to the Mediterranean diet covers the main alternative tested in Keto-Med.

Where keto is a medical treatment

There is one area where ketogenic diets are an established clinical therapy: drug-resistant epilepsy. A 2020 Cochrane review of 13 studies with 932 participants found ketogenic diets increased seizure freedom and seizure reduction in children compared with usual care, while the evidence in adults "remains uncertain" (Martin-McGill and colleagues, Cochrane 2020). The certainty of evidence was rated low to very low, mainly because the studies were small and unblinded. Medical ketogenic diets are run by specialist teams with careful monitoring, which is very different from following a keto plan from social media.

What the evidence says about carnivore

There are no published randomised controlled trials of the carnivore diet. The largest study is a 2021 social media survey of 2,029 adults who said they had eaten a carnivore diet for at least six months (Lennerz and colleagues, Current Developments in Nutrition 2021). Respondents reported high satisfaction, few side effects and lower body mass index. Among those who reported blood results, LDL cholesterol was "markedly elevated" at a median of 172 mg/dL (about 4.4 mmol/L), while HDL and triglycerides looked favourable.

This study deserves credit for collecting data where almost none existed, but it cannot show that carnivore is safe or effective:

• People were recruited from carnivore communities online, so those who struggled or stopped are unlikely to have taken part.
• All results were self-reported, including weight and blood tests.
• There was no comparison group.
• The authors themselves concluded that "the long-term effects of this dietary pattern require further study".

The biggest open questions are the long-term effects of very high LDL cholesterol, eating no fibre, and eating large amounts of red and processed meat. The World Health Organization's cancer agency classifies processed meat as Group 1, "carcinogenic to humans", and red meat as Group 2A, "probably carcinogenic" (WHO). The NHS advises people who eat more than 90 g of red or processed meat a day to cut down to 70 g (NHS, meat in your diet). A carnivore day can easily contain several times that.

Fibre, cholesterol and nutrients: the shared risks

The further a diet moves from plant foods, the larger these gaps become.

Fibre

UK guidance says adults should eat 30 g of fibre a day, and most adults already average only about 20 g (NHS). Keto removes grains, pulses and most fruit, the main fibre sources, so intakes usually fall further; Keto-Med measured lower fibre on keto than on the Mediterranean diet. Carnivore provides close to none. Our fibre calculator shows how far your current day falls short.

LDL cholesterol

Higher LDL cholesterol appeared on keto in the 2013 meta-analysis, in Keto-Med, and in the carnivore survey. The size of the rise varies hugely between individuals, so anyone following these diets long term should have their lipids checked, ideally before starting and again after a few months.

Micronutrients

Cutting whole food groups narrows the range of nutrients. Removing dairy, as paleo does, makes calcium harder to reach. Removing fruit, pulses and whole grains, as keto and carnivore do, lowers folate, potassium, magnesium and vitamin C. The answer is not automatically supplements; a dietitian can check what your version of the diet actually supplies.

What did early humans actually eat?

Paleo rests on a historical claim, and the archaeological evidence complicates it.

Hunter-gatherer diets were not one diet. They varied by place, season and era, and many relied heavily on plants. Starch grains on grinding stones from sites in Italy, Russia and the Czech Republic show that processing of wild plants, possibly into flour, was "a common practice, widespread across Europe from at least ~30,000 y ago" (Revedin and colleagues, PNAS 2010). Isotope analysis of Later Stone Age hunter-gatherers at Taforalt in Morocco, from 15,000 to 13,000 years ago, found "a substantial plant-based component" in their diets, challenging the idea that pre-agricultural people relied mainly on animal protein (Moubtahij and colleagues, Nature Ecology and Evolution 2024).

None of this means early humans ate like a modern guideline. It means there is no single "Paleolithic diet" to copy, and that the meat-heavy picture used to sell paleo and carnivore is a simplification. More importantly, what people ate tens of thousands of years ago tells us little about what keeps people healthy into their eighties today. That question is better answered by trials and long-term studies in modern populations.

Social media trends versus trial evidence

All three diets spread largely through social media, books and podcasts, and personal transformation stories are compelling. They are also exactly the kind of evidence that misleads most often.

• Survivorship. You hear from people for whom the diet worked, not from the many who quietly stopped.
• Short timelines. Most diets produce early weight loss, partly from water and glycogen when carbohydrate falls. Trials show that leads tend to shrink by 12 to 24 months.
• Several changes at once. Someone who starts carnivore often also stops drinking alcohol, cuts ultra-processed snacks and starts training. The diet gets the credit for all of it.
• Biomarker selection. Lower triglycerides and higher HDL are real, but they are often reported without the LDL rise beside them.

The same critical lens applies to other popular plans; our guides to detox diets and the cortisol diet look at two more where the claims run well ahead of the evidence. A quick way to see how your current eating compares with an evidence-based pattern is our free healthy eating check.

Take the free healthy eating check

Restrictive diets and disordered eating

Any diet built on strict rules and forbidden foods carries a risk for some people. In an Australian cohort of nearly 1,700 adolescents followed for three years, girls who dieted severely were 18 times more likely to develop an eating disorder than girls who did not diet, and the authors concluded that dieting was "the most important predictor of new eating disorders" (Patton and colleagues, BMJ 1999). That study looked at dieting in general among teenagers, not at these three diets, but the mechanism, rigid restriction and moralising food, is the same.

Warning signs that a diet is becoming a problem include anxiety when "allowed" foods are unavailable, avoiding social meals, cutting out more and more foods, and feeling guilt after eating. If you have a current or past eating disorder, avoid highly restrictive diets unless your treatment team supports it.

Who should avoid these diets or get supervision

Talk to a doctor or registered dietitian before starting keto or carnivore, and think carefully about paleo, if you:

• have type 1 or type 2 diabetes, especially if you take insulin, sulfonylureas or SGLT2 inhibitors;
• are pregnant, breastfeeding or planning a pregnancy;
• have kidney disease, where high protein intakes and electrolyte shifts matter;
• have high cholesterol, familial hypercholesterolaemia or heart disease;
• have a history of an eating disorder;
• are a child or teenager, unless the diet is a supervised medical treatment;
• take medicines for blood pressure, which may need adjusting as weight and fluid balance change.

A practical takeaway

If you are drawn to one of these diets, it helps to separate the useful core from the extremes:

1. Keep what the evidence supports. Less sugar, fewer refined grains and ultra-processed foods, more vegetables, and protein at each meal. These are features of paleo and well-formulated keto that also appear in the Mediterranean diet and other well-studied patterns, including the anti-inflammatory diet.
2. Question the exclusions. Whole grains, pulses, fruit and dairy each have good evidence behind them. Remove them only for a medical reason.
3. Check your energy needs. Weight loss on any diet comes down mostly to eating less overall. Our calorie calculator gives you a starting estimate.
4. Measure, do not assume. If you follow keto or carnivore for more than a few weeks, ask for a lipid test.
5. Pick what you can keep doing. DIETFITS and Keto-Med both suggest the best diet is the healthy one you can sustain.

If you would like a plan that fits your health, medicines and preferences, you can book a session with a registered dietitian. Our explainer on the new food pyramid shows how current guidance frames a balanced diet.