evipedia.ai
← Back

Paleo Diet for Health & Longevity

Evidence Review created on 10/25/2026 using AI4L / Opus 5.5

Also known as: Paleolithic Diet, Palaeolithic Diet, Paleo, Caveman Diet, Stone Age Diet, Hunter-Gatherer Diet

Motivation

The paleo diet (Paleolithic diet) is an eating pattern built around foods presumed to have been available to human hunter-gatherers before farming began: meat, fish, eggs, vegetables, fruit, nuts and seeds. It leaves out grains, beans and lentils, dairy foods, added sugar, added salt and industrially processed foods. Its appeal to people focused on long-term health is the promise of a whole-food diet matched to human evolutionary history, with possible relevance to body weight, heart health and blood sugar control.

The idea grew from a medical journal essay proposing that the chronic diseases of modern societies reflect a mismatch between ancient genes and recently introduced foods. Popular books turned it into a broad movement, and research groups in Sweden, the United States, Australia and the Netherlands began testing it in controlled studies, mostly in people with excess weight, type 2 diabetes or heart disease. Critics question both its picture of ancestral eating and whether its exclusions are needed.

This review examines the human evidence on the benefits and nutritional risks of the paleo diet, its practical demands, and the open questions that matter for adults willing to follow a demanding eating plan over the long term.

Benefits - Risks - Protocol - Conclusion

Expert commentary, scientific reviews and a lecture giving a high-level overview of the paleo diet’s rationale, evidence and criticisms.

Note on priority experts: Rhonda Patrick, Peter Attia, Andrew Huberman, Life Extension and Lifespan.io mention the paleo diet only in passing (for example, Attia lists it as one example of dietary restriction, and Patrick’s interview with Terry Wahls centres on micronutrients rather than the diet); none discusses it in substantial depth, so none is listed, and only four items qualify.

Grokipedia

Paleolithic diet

An AI-generated encyclopedia overview of the diet’s definition, history, evolutionary rationale, trial evidence and criticisms, useful as a broad starting summary.

Examine

Paleolithic Diet

An independent evidence summary concluding that well-formulated paleo diets perform comparably to other healthy diets for weight and metabolic markers, while restrictive versions risk nutrient deficiencies.

ConsumerLab

No dedicated ConsumerLab article on the paleo diet exists; the site’s search returns only product reviews that mention the term and a brief clinical update on paleo eating and iodine within its multivitamin review.

Systematic Reviews

Meta-analyses (statistical pooling of studies) of randomized trials and cohort studies on the paleo diet’s effects on body weight, cardiometabolic markers and chronic disease, plus a systematic review of micronutrient shortfalls.

No systematic review or meta-analysis has assessed bone density, fractures or iodine-related thyroid harm on the diet; the risk side is represented only by micronutrient intake and status data.

Mechanism of Action

The diet acts through the foods it removes and those it concentrates. As a dietary pattern, it has no half-life, selectivity or metabolic pathway.

  • Fewer calories through fullness: more protein and produce raise satiety per calorie; in a Lund trial participants felt equally full while eating fewer calories than on a Mediterranean-like diet (Jönsson et al., 2010). Weight loss then improves insulin sensitivity (how strongly cells respond to insulin), liver fat and blood fats.
  • Lower glycemic load: (how much the diet raises blood glucose) removing grains and sugar flattens glucose and insulin spikes.
  • Sodium–potassium shift: no added salt and abundant produce lower sodium and raise potassium, which lowers blood pressure.
  • Fat quality: fish, nuts and olive oil raise monounsaturated and polyunsaturated fats (MUFA and PUFA, fats with one or several double bonds), proposed to drive liver-fat loss (Otten et al., 2016).
  • Removal of ultra-processed foods: (industrial products of refined ingredients and additives) in a Spanish cohort, the paleo score’s link with lower heart disease weakened when this item was removed (de la O et al., 2022).
  • Competing explanations: proponents argue that grain and legume lectins (carbohydrate-binding plant proteins) increase intestinal permeability (gut-lining leakiness) and inflammation; a two-week trial found no permeability change (Boers et al., 2014), and a weight-stable trial matched for macronutrients (protein, fat and carbohydrate) found no glycemic difference versus a diabetes diet (Fontes-Villalba et al., 2025), suggesting calorie reduction and food quality, not the exclusions themselves, drive most effects.

Historical Context & Evolution

The paleo diet began not as a treatment but as an anthropological model. In 1985, physician S. Boyd Eaton and anthropologist Melvin Konner argued that human genes were shaped over millions of years of foraging, that ancestral diets varied widely by period, place and season, and that modern diet-related diseases reflect a discordance (mismatch) between those genes and recent foods (Eaton & Konner, 1985). Loren Cordain extended the idea in an influential review naming seven altered dietary traits, including glycemic load, fatty acid balance, sodium–potassium ratio and fiber (Cordain et al., 2005) and in popular books sold under The Paleo Diet brand, a commercial interest in the diet’s adoption.

The shift toward health optimization came when Swedish researchers tested the model in randomized trials, beginning with people with heart disease and glucose intolerance (Lindeberg et al., 2007). Fitness communities and books then popularized it widely.

Archaeological evidence complicated the historical premise: starch grains on 30,000-year-old grinding stones show plant processing, possibly flour making (Revedin et al., 2010), and Neanderthal dental plaque contained grass seeds, some cooked, and legumes (Henry et al., 2011). Proponents respond that the model has been revised in light of such findings (Konner & Eaton, 2023). Opinion has since moved from debating ancestral accuracy to testing outcomes directly, with trials suggesting short-term metabolic benefits while long-term and hard-outcome data remain sparse.

Expected Benefits

High 🟩 🟩 🟩

Body Weight and Waist Reduction ⚠️ Conflicted

Paleo diets produced more weight and waist loss than guideline-based diets, likely through greater fullness per calorie. Randomized controlled trials (RCTs, studies assigning participants by chance) from independent groups led by Lindeberg (Lindeberg et al., 2007), Mellberg (Umeå), Boers (Boers et al., 2014) and Genoni (Genoni et al., 2016) agree, and an 11-trial meta-analysis confirms the effect. In the longest trial, the fat-loss advantage was no longer significant at 24 months (Mellberg et al., 2014). Net reading: a replicated short-term advantage whose durability depends on sustained adherence.

Magnitude: 3.52 kg more weight loss than guideline-based control diets (95% CI, confidence interval, the range likely to contain the true value: 1.79 to 5.26 kg), pooled from 11 RCTs (de Menezes et al., 2019); at 24 months fat-mass loss was 4.6 vs 2.9 kg (within-group changes, difference not reported; between-group P=0.095; P is the probability of a difference this large arising by chance) (Mellberg et al., 2014).

Lower Triglycerides and Blood Lipids ⚠️ Conflicted

Paleo diets lowered LDL cholesterol (low-density lipoprotein, artery-depositing cholesterol), a validated surrogate, and triglycerides, likely via less refined carbohydrate and more unsaturated fat. Triglyceride reductions were replicated by Jönsson (Jönsson et al., 2009), Mellberg (Mellberg et al., 2014) and Boers (Boers et al., 2014), persisting 24 months in Mellberg’s trial; pooled trials from these and other groups show LDL falls (Sohouli et al., 2022). A four-week trial in healthy women found no difference (Genoni et al., 2016). Net reading: consistent lipid benefit with metabolic risk, little evidence in healthy adults.

Magnitude: LDL cholesterol 0.35 mmol/L lower than control diets (95% CI 0.03 to 0.67 mmol/L lower), 10 RCTs in metabolic disorders (Sohouli et al., 2022); triglycerides 0.40 mmol/L lower than guideline-based diets (95% CI 0.04 to 0.76 mmol/L lower) at up to 6 months, 4 RCTs (Manheimer et al., 2015).

Lower Blood Pressure ⚠️ Conflicted

Removing added salt and adding potassium-rich produce lowers blood pressure; reductions versus control diets came from trials led by Boers (Netherlands; Boers et al., 2014) and Jönsson (Lund; Jönsson et al., 2009). A ten-trial meta-analysis found significant systolic and diastolic reductions, whereas an earlier four-trial meta-analysis found reductions whose confidence intervals crossed zero (Manheimer et al., 2015). Part of the effect tracks weight loss. Net reading: a modest, probably real reduction, shown mainly in people with metabolic disorders.

Magnitude: Systolic 5.89 mmHg lower (95% CI 1.86 to 9.97) and diastolic 4.01 mmHg lower (95% CI 1.80 to 6.21) than control diets (Sohouli et al., 2022); earlier pooled systolic estimate 3.64 mmHg lower (95% CI 7.36 lower to 0.08 higher) (Manheimer et al., 2015).

Medium 🟩 🟩

Reduced Liver Fat ⚠️ Conflicted

In obese postmenopausal women, an ad libitum (unrestricted-quantity) paleo diet reduced liver fat, measured by magnetic resonance spectroscopy (scanner-based fat quantification), more than a low-fat diet at 6 months, independent of weight change, possibly through its higher unsaturated fat (Otten et al., 2016). In the same trial the between-group difference disappeared by 24 months. The evidence comes from a single trial by one Umeå research group. Net reading: a short-term advantage over low-fat eating that did not persist to two years.

Magnitude: Liver fat fell 64% (95% CI 54–74%) on the paleo diet vs 43% (95% CI 27–59%) on the low-fat diet at 6 months (within-group changes; between-group P<0.01); 50% vs 49% at 24 months (Otten et al., 2016).

Low 🟩

Improved Glycemic Control and Insulin Sensitivity ⚠️ Conflicted

Glucose tolerance improved more than on a Mediterranean-like diet in ischemic heart disease (reduced heart blood flow) (Lindeberg et al., 2007) and HbA1c more than on a diabetes diet (Jönsson et al., 2009); a four-trial meta-analysis found no difference (Jamka et al., 2020). Net reading: inconsistent, mostly alongside weight loss.

Magnitude: HbA1c 0.4 percentage points lower than the diabetes diet (P=0.01; no CI reported) in a 13-person crossover (each person ate both diets) (Jönsson et al., 2009); versus healthy diets, pooled SMD (standardized mean difference, effect in standard-deviation units) −0.38 (95% CI −0.87 to 0.11), not significant (Jamka et al., 2020).

Lower Mortality and Cardiovascular Disease ⚠️ Conflicted

Cohorts scoring diets for paleo resemblance, not people following it, linked high scores to lower all-cause and cancer mortality (Whalen et al., 2017) and fewer cardiovascular events (de la O et al., 2022); an Iowa cohort found no association (Cheng et al., 2018). Net reading: suggestive but indirect and inconsistent.

Magnitude: All-cause mortality HR (hazard ratio, relative event rate over time) 0.77 (95% CI 0.67–0.89), highest vs lowest score quintile (fifth of participants ranked by score) (Whalen et al., 2017); pooled RR (relative risk) 0.90 (95% CI 0.87–0.94) for all-cause and 0.90 (95% CI 0.85–0.97) for cancer mortality, and 0.84 (95% CI 0.70–1.00) for coronary heart disease incidence, across cohorts (Bahrami et al., 2026); cardiovascular disease HR 0.45 (95% CI 0.27–0.76), highest vs lowest quintile, with 28 cases per 33,985 person-years (follow-up years summed across all participants) vs 42 per 55,306 (de la O et al., 2022).

Reduced Fatigue and Better Quality of Life in Multiple Sclerosis

In multiple sclerosis (MS), fatigue and quality of life improved within the Wahls modified paleo group (no usual-diet control); Terry Wahls sells books and courses on it (Wahls et al., 2021); her web course cut fatigue versus waitlist (Titcomb et al., 2023). Evidence is indirect (modified versions, supplements, lifestyle add-ons).

Magnitude: Fatigue SMD −1.27 (95% CI −1.81 to −0.74), physical quality of life SMD 1.01 (95% CI 0.40 to 1.63) and mental quality of life SMD 0.81 (95% CI 0.26 to 1.37) vs usual diet in a network meta-analysis (pooling direct and indirect trial comparisons), very low credibility (Snetselaar et al., 2023).

Improved Autoimmune Thyroid Markers ⚠️ Conflicted

A mostly case-report review, co-authored by a True Paleo Inc. affiliate, linked paleo-type diets to improved thyroid antibodies in Hashimoto’s thyroiditis (autoimmune thyroid inflammation) (Hollywood et al., 2023). Evidence is indirect (coached, stricter elimination versions); an uncontrolled pilot found no antibody change (Abbott et al., 2019). Net reading: benefit unconfirmed.

Magnitude: Symptom questionnaire score fell from 92 to 29 over 10 weeks, with no significant change in thyroid antibodies or hormones (no control group) (Abbott et al., 2019).

Speculative 🟨

Lower Systemic Inflammation ⚠️ Conflicted

Pooled trials show lower C-reactive protein (Sohouli et al., 2022); a two-week trial found no inflammation change (Boers et al., 2014). This marker is unvalidated as an outcome surrogate. Net reading: possible, unconfirmed.

Lower Dietary Acid Load

A 14-day paleo diet lowered measured net acid excretion (acid removed by the kidneys) versus a diabetes diet (Frassetto et al., 2013). Proposed bone and kidney benefits are mechanistic only and untested.

Benefit-Modifying Factors

  • Genetic polymorphisms: No paleo trial has stratified by genotype. APOE4 (a variant of the gene for a cholesterol-transport protein) carriers showed larger triglyceride and C-reactive protein responses to dietary fat changes (Carvalho-Wells et al., 2012), so fat composition may matter more for them.
  • Baseline biomarkers: The most insulin-resistant participants improved most in insulin sensitivity (Masharani et al., 2015); lipid and blood-pressure gains appear in people with metabolic syndrome but not in healthy women (Genoni et al., 2016).
  • Sex: Long-term data come mainly from postmenopausal women (Mellberg et al., 2014); a short-term trial in men with heart disease showed similar metabolic gains (Lindeberg et al., 2007). No trial has formally compared benefit by sex.
  • Pre-existing conditions: Benefits are documented mainly in obesity, metabolic syndrome, type 2 diabetes and ischemic heart disease; in MS, a modified version targets fatigue. Benefit in already lean, metabolically healthy adults is largely untested.
  • Age: Trials enrolled adults mostly aged 45–75; for those at the older end, weight-loss benefits come alongside lean-mass and bone losses, which make protein adequacy and resistance training more important.
  • Starting diet: Gains are largest when the diet replaces a highly processed diet; the cohort association with heart disease depended largely on avoiding ultra-processed foods (de la O et al., 2022).

Potential Risks & Side Effects

High 🟥 🟥 🟥

No risk reaches High: the documented harms are nutrient-status, nutrient-intake and body-composition changes from single trials or food records, and no trial has measured fractures, thyroid disease or cancer.

Medium 🟥 🟥

Iodine Deficiency

Strict versions exclude iodized salt and dairy, major iodine sources in many countries. In a 2-year Swedish trial, urinary iodine halved on the paleo diet but not on the Nordic Nutrition Recommendations diet, and free triiodothyronine (FT3, active thyroid hormone) was lower at 6 months (Manousou et al., 2018). An Australian trial found lower iodine intake (Genoni et al., 2016). Stakes are highest in pregnancy and breastfeeding, when iodine supports fetal brain development.

Magnitude: Iodine intake 47.9 µg/day lower than the Australian Guide to Healthy Eating diet (95% CI 16.5 to 79.2 µg/day lower) over 4 weeks (Genoni et al., 2016).

Low 🟥

Lean Mass Loss Without Resistance Training

In a 12-week type 2 diabetes trial, men on the paleo diet with standard exercise advice lost more lean mass than men also doing supervised training three times weekly (Otten et al., 2017). Evidence is indirect: no non-paleo arm, so the loss is not shown to be paleo-specific.

Magnitude: Men lost 2.6 kg lean mass on the paleo diet alone vs 1.2 kg with supervised exercise (within-group changes; between-group P<0.05); the source has no non-paleo comparison group (Otten et al., 2017).

Low Calcium Intake and Potential Bone Loss

Excluding dairy lowered calcium intake versus guideline diets in Australian (Genoni et al., 2016) and Swedish (Jönsson et al., 2009) trials. Evidence is indirect: no trial measured bone density or fractures.

Magnitude: Calcium intake 292 mg/day lower than the Australian Guide to Healthy Eating diet (95% CI 99 to 486 mg/day lower) over 4 weeks (Genoni et al., 2016).

Ketoacidosis During Breastfeeding on Very-Low-Carbohydrate Versions

Breastfeeding raises glucose demand, and very-low-carbohydrate eating can then trigger ketoacidosis (dangerous blood acid build-up) requiring hospital care (von Geijer & Ekelund, 2015). Examine lists this risk for strict paleo versions. Evidence is indirect: isolated case reports, mostly on other low-carbohydrate diets.

Magnitude: Not quantified in available studies. Only isolated case reports exist, so no incidence or comparison-group rate is available.

Colorectal Cancer Risk From Red Meat ⚠️ Conflicted

Red meat is linked to colorectal cancer (Chan et al., 2011); evidence is indirect because paleo trials never measured cancer. A cohort scoring paleo-like diets found lower colorectal cancer risk (Xiao et al., 2023). Net reading: risk likely depends on red-meat amount.

Magnitude: RR 1.17 per 100 g/day of fresh red meat (95% CI 1.05–1.31) in a meta-analysis of cohorts (Chan et al., 2011); paleo-score cohort HR 0.76 (95% CI 0.61–0.95), highest vs lowest quartile (quarter ranked by score), with 0.84 vs 1.27 cases per 1,000 person-years (Xiao et al., 2023).

Speculative 🟨

Less Resistant Starch and Higher TMAO (Trimethylamine N-oxide) ⚠️ Conflicted

Strict followers ate less resistant starch (fiber-like starch); TMAO (a heart-disease-linked metabolite) was higher (Genoni et al., 2020); a 4-week trial showed lower resistant starch, no TMAO rise (Genoni et al., 2019). Net reading: unresolved.

Kidney Stone Formation

High-protein diets raise urinary calcium and uric acid, while fruit and vegetables raise protective citrate (Barghouthy et al., 2021). No paleo trial has measured stones; the basis is mechanistic.

Risk-Modifying Factors

  • Genetic polymorphisms: APOE4 carriers showed C-reactive protein rises on high-saturated-fat diets (Carvalho-Wells et al., 2012), relevant to butter- and coconut-heavy versions. HFE variants (a gene regulating iron absorption) causing hemochromatosis (iron overload) may be worsened by heme-iron-rich red meat (theoretical).
  • Baseline biomarkers: Low starting iodine status raises deficiency risk; trial participants were already below the population iodine adequacy cutoff at baseline (Manousou et al., 2018). Low bone density or high LDL cholesterol raises the stakes of calcium shortfall or saturated-fat-heavy menus.
  • Sex: Women of reproductive age face the iodine risk most acutely; postmenopausal women face the calcium and bone risk. Men lost more lean mass without supervised training in the only trial reporting it (Otten et al., 2017).
  • Pre-existing conditions: Chronic kidney disease (CKD) raises concern about high potassium (Palmer & Clegg, 2016) and protein intake (per KDIGO, Kidney Disease: Improving Global Outcomes, a nonprofit without diet-advice revenue; KDIGO, 2024); thyroid disease, kidney stones, osteoporosis and past eating disorders also heighten risk.
  • Age: Older adults have higher calcium needs and greater vulnerability to lean-mass and bone loss during weight reduction; kidney function also declines with age, raising the relevance of protein and potassium load.

Key Interactions & Contraindications

  • Warfarin, other vitamin K antagonists (vitamin K-blocking clot-preventing drugs; acenocoumarol): Leafy greens add vitamin K; in a randomized crossover, vitamin K-rich diets lowered INR (international normalized ratio, a clotting-time measure) (Franco et al., 2004). Consequence: weaker anticoagulation; mitigation: steady intake, INR checks after switching.
  • Lithium: Eliminating added salt lowers sodium intake; low sodium reduced lithium clearance in volunteers (Folkerd et al., 1995). Consequence: rising lithium levels and toxicity; mitigation: serum lithium check within 1–2 weeks of starting.
  • Insulin and sulfonylureas (glimepiride, glipizide; drugs that make the pancreas release insulin) (theoretical): Lower carbohydrate intake and weight loss reduce glucose. Consequence: hypoglycemia (low blood sugar); mitigation: frequent glucose checks and prescriber-led dose review.
  • Antihypertensives (blood-pressure-lowering drugs; lisinopril, amlodipine, hydrochlorothiazide) (theoretical): Lower sodium, higher potassium and weight loss add to drug effects. Consequence: hypotension (low blood pressure) and dizziness; mitigation: home blood-pressure monitoring and dose review.
  • Potassium-raising drugs: ACE inhibitors (block angiotensin formation; lisinopril), ARBs (angiotensin receptor blockers; losartan), spironolactone (theoretical): High-potassium produce adds to drug-related potassium retention, especially with impaired kidneys (Palmer & Clegg, 2016). Consequence: hyperkalemia (high blood potassium); mitigation: potassium monitoring.
  • Over-the-counter NSAIDs (nonsteroidal anti-inflammatory drugs; ibuprofen, naproxen) and potassium-chloride salt substitutes (theoretical): Both raise blood potassium in impaired kidneys, adding to a potassium-rich diet. Consequence: hyperkalemia; mitigation: limit regular use and check potassium.
  • Multi-supplement regimens: In a modified paleo program, supplements pushed several participants above upper limits for niacin, vitamin D, magnesium and zinc (Titcomb et al., 2020). Consequence: nutrient toxicity; mitigation: total-intake review including food.
  • Additive supplements (theoretical): Glucose-lowering supplements (berberine, chromium) and blood-pressure-lowering supplements (garlic extract, potassium) add to the diet’s effects. Consequence: hypoglycemia, hypotension or hyperkalemia; mitigation: introduce one at a time with monitoring.
  • GLP-1 receptor agonists (semaglutide, tirzepatide; appetite-reducing gut-hormone mimics) (theoretical): Combined appetite suppression with a protein-dependent diet may produce large calorie deficits. Consequence: accelerated lean-mass loss; mitigation: protein targets and resistance training.
  • Exercise: Supervised training added fitness and lean-mass preservation without blunting metabolic gains (Otten et al., 2017). Consequence: complementary effects; mitigation: none needed beyond fueling adequately.

Populations who should avoid Paleo Diet:

  • People with non-dialysis CKD stages G3–G5 (eGFR, estimated glomerular filtration rate, a measure of kidney filtering, below 60 mL/min/1.73 m²): the diet’s protein load (about 30% of energy) exceeds the 0.8 g/kg/day suggested by KDIGO, a nonprofit earning no revenue from diet advice (KDIGO, 2024).
  • Pregnant or breastfeeding women not taking an iodine supplement, given the documented fall in iodine status (Manousou et al., 2018).
  • People with a current or past eating disorder, for whom strict food-group exclusion can reinforce restrictive eating.
  • People with osteoporosis who cannot replace dairy calcium from other foods or supplements (Pitt, 2016).

Risk Mitigation Strategies

Doses and timings below follow common practice unless cited.

  • Iodine supplementation: Iodized salt or a 150 µg/day iodine supplement offsets the iodine deficiency documented on strict versions; the trial authors suggest supplementation (Manousou et al., 2018).
  • Iodine threshold: Rising thyroid-stimulating hormone (a pituitary signal to the thyroid) prompts clinical review for iodine deficiency; WHO’s (World Health Organization; its members earn no revenue from it) 100 µg/L urinary iodine cutoff applies to population medians only (WHO, 2013).
  • Calcium replacement: Intake of 1,000–1,200 mg/day from canned fish with bones, leafy greens, almonds, mineral water or supplements offsets the calcium shortfall and protects bone.
  • Resistance training: Two to three sessions weekly, as in the supervised arm (three sessions weekly) that limited lean-mass loss in men (Otten et al., 2017), counter muscle loss during weight reduction.
  • Protein moderation in kidney disease: Protein near 0.8 g/kg/day in CKD stages G3–G5, per KDIGO, a nonprofit without revenue from diet advice (KDIGO, 2024), avoids accelerating kidney decline.
  • Fish and poultry over red meat: Limiting red meat (about three servings weekly) and avoiding processed meat addresses the colorectal cancer association seen with higher red and processed meat intake in cohort studies (Chan et al., 2011).
  • Fiber diversity: Cooled cooked potatoes, green bananas and diverse vegetables restore resistant starch, addressing the gut microbiota and TMAO concern.
  • Medication review before starting: Prescriber review of insulin, sulfonylureas, antihypertensives, warfarin and lithium within the first 1–2 weeks prevents hypoglycemia, hypotension, INR drift or lithium toxicity.
  • Potassium threshold: Medical review is indicated if serum potassium exceeds the laboratory reference range, which signals hyperkalemia risk in people with CKD or on potassium-raising drugs.

Therapeutic Protocol

Timing, cycling, titration steps and other non-dose parameters below reflect common practice unless cited.

  • Core template (Cordain): Lean meat, fish, eggs, vegetables, fruit and nuts; excludes grains, legumes, dairy, added sugar, salt and refined oils. Popularized by Loren Cordain (Cordain et al., 2005).
  • Research macronutrient regimen (Umeå): Ad libitum, 30% of energy from protein, 40% from fat (emphasizing MUFA and PUFA) and 30% from carbohydrate, with dairy, cereals, added salt and refined fats excluded (Mellberg et al., 2014).
  • Lund trial regimen: Ad libitum lean meat, fish, fruit, vegetables, root vegetables, eggs and nuts, without calorie targets (Lindeberg et al., 2007).
  • Modified paleo for MS (Wahls): Nine or more cups of fruit and vegetables daily, grains and dairy excluded (Chenard et al., 2019); developed by Terry Wahls, who sells the protocol’s books and courses.
  • Flexible template (Kresser): Treats paleo as an adjustable template rather than fixed rules, leaving room for individually tolerated foods such as some dairy (Kresser, Beyond Paleo).
  • Autoimmune protocol (AIP): Popularized by Sarah Ballantyne; additionally removes eggs, nuts, seeds and nightshades (tomatoes, peppers, eggplant); tested as 6 weeks of elimination then 5 weeks of maintenance in inflammatory bowel disease (Konijeti et al., 2017).
  • Time of day: No trial has tested meal timing; regular meals with protein at each meal are common practice, and pairing with an earlier last meal is a frequent addition.
  • Genetic polymorphisms: APOE4 carriers may favor unsaturated over saturated fats (Carvalho-Wells et al., 2012); HFE hemochromatosis carriers may limit red meat. No pharmacogenetic variants apply.
  • Sex: Iodine intake is most relevant for women of reproductive age, resistance training for men’s lean mass (Otten et al., 2017), and calcium intake for postmenopausal women.
  • Age: In older adults, protein adequacy, calcium intake, resistance training and kidney function carry more weight, particularly with high-protein versions.
  • Baseline biomarkers: The most insulin-resistant participants showed the largest gains in insulin sensitivity (Masharani et al., 2015); low baseline iodine or bone density makes supplementation more relevant.
  • Pre-existing conditions: In diabetes and hypertension, medication doses interact with the diet’s effects; in CKD, KDIGO, a nonprofit without diet-advice revenue, advises protein moderation (KDIGO, 2024); MS and inflammatory bowel disease versions follow the Wahls or AIP templates.

Discontinuation & Cycling

  • Lifelong versus short-term: Intended as a permanent pattern; benefits fade as adherence falls, and only 35% of self-selecting followers continued at 12 months (Jospe et al., 2020).
  • Withdrawal effects: No physiological withdrawal occurs on stopping. Transitional headache or fatigue in the first week of carbohydrate reduction is anecdotal; weight regain is the main consequence of stopping.
  • Tapering: Not required; stepwise reintroduction of one excluded food group at a time commonly identifies poorly tolerated foods; the AIP trial kept all exclusions through maintenance, offering staged reintroduction afterward (Konijeti et al., 2017).
  • Cycling: No evidence supports cycling for efficacy; a strict 30-day phase followed by a flexible template is a common practitioner approach.
  • Medication re-review on stopping: If diabetes or blood-pressure drugs were reduced, glucose and blood pressure may rise after stopping, requiring renewed monitoring.

Sourcing and Quality

  • Meat quality: Grass-fed beef has more omega-3 fats and antioxidant precursors than grain-fed beef, though higher-fat grain-fed cuts can supply similar amounts (Daley et al., 2010).
  • Processed versus fresh meat: Bacon, sausages and cured meats carry the strongest colorectal cancer association (Chan et al., 2011); fresh fish, poultry and unprocessed cuts fit the diet’s whole-food intent.
  • Fish selection: Low-mercury, oily species such as salmon, sardines and mackerel supply omega-3 fats with less mercury than swordfish, king mackerel or bigeye tuna.
  • Packaged “paleo” products: Grain-free snacks and baked goods can be ultra-processed; “Certified Paleo” marks from the Paleo Foundation, a private certifier, indicate ingredient compliance, not nutritional quality.
  • Iodine sources: Strict versions drop iodized salt; seafood, eggs and iodized salt are reliable sources, while seaweed iodine content varies widely between products.
  • Supplements: Where iodine, calcium or vitamin D supplements are used, third-party tested products (e.g., verified by USP, the United States Pharmacopeia, or NSF, a certification body) reduce contamination and dosing errors.

Practical Considerations

  • Time to effect: Blood pressure, triglycerides and cholesterol changed within 2 weeks (Boers et al., 2014); glucose tolerance within 12 weeks (Lindeberg et al., 2007); liver fat by 6 months (Otten et al., 2016).
  • Adherence: Participants found the diet satiating but difficult to adhere to (Jönsson et al., 2013); in the 2-year trial, protein targets were poorly met (Mellberg et al., 2014).
  • Common pitfalls: Relying on red and processed meat, eating grain-free baked treats, neglecting iodine and calcium, cutting starch so low that training suffers, and adopting an all-or-nothing strictness that ends in dropout.
  • Regulatory status: A whole-food eating pattern requiring no prescription or approval; “Certified Paleo” marks on packaged foods come from private certifiers such as the Paleo Foundation.
  • Cost and accessibility: Meat, fish, nuts and fresh produce cost more than grain-based staples; the diet is accessible but not inexpensive, and eating out requires planning.

Interaction with Foundational Habits

  • Sleep: No direct interaction documented; no paleo trial has measured sleep. Indirectly, weight loss may reduce obstructive sleep apnea (breathing pauses during sleep) severity, and very-low-carbohydrate dinners may affect sleep onset in some individuals.
  • Nutrition: Direct, since the diet is itself a nutrition intervention. It depletes iodine, calcium and resistant starch (Genoni et al., 2020) and raises potassium and vitamin intake; practical additions include iodized salt, bony fish, leafy greens and cooled cooked potatoes.
  • Exercise: Potentiating: supervised aerobic and resistance training on a paleo diet preserved lean mass in men and raised fitness (Otten et al., 2017) and reduced heart-muscle fat (Otten et al., 2019). Trials in athletes are lacking, so effects on performance remain unknown (Frączek et al., 2021).
  • Stress management: No direct interaction found: a two-week trial found no change in salivary cortisol (a stress hormone) (Boers et al., 2014). Indirectly, rigid food rules and social eating constraints can add stress; a flexible template may reduce it.

Monitoring Protocol & Defining Success

Before starting, a baseline assessment documents body weight, waist circumference and blood pressure, plus fasting blood tests covering triglycerides, LDL cholesterol, HbA1c, glucose and thyroid-stimulating hormone. People with kidney disease or potassium-raising drugs add potassium and eGFR; people on warfarin add INR; postmenopausal women and older men at bone risk add a bone density scan. A urinary iodine test documents starting iodine status, which matters most for women who may become pregnant.

Ongoing monitoring follows a cadence of a check at 6–12 weeks, then every 6–12 months while the diet continues, with bone density repeated every 2 years where measured. Closer checks in the first 2 weeks are common practice for people on insulin, sulfonylureas, lithium or warfarin. Success means sustained improvement in waist, triglycerides and glucose markers without iodine, calcium or lean-mass shortfalls.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Body weight and waist circumference No established target: track change from own baseline Expected to change (fall) Measure fasting, same time of day; waist at navel level
Fasting triglycerides No established target: track change from own baseline Expected to change (fall) 10–12-hour fast; pair with full lipid panel
LDL cholesterol No established target: track change from own baseline Safety check: a rise prompts a change in fat sources Fasting or non-fasting; pair with apolipoprotein B (ApoB, a count of atherogenic particles) if LDL rises
HbA1c and fasting glucose HbA1c 4.0–5.6%; fasting glucose 70–99 mg/dL (standard reference range) Expected to change in insulin resistance HbA1c reflects about 3 months; adjust diabetes drugs with prescriber
Thyroid-stimulating hormone (TSH) 0.4–4.0 mIU/L (standard reference range) Safety check for iodine deficiency Morning draw; pair with free thyroxine (FT4, the main hormone the thyroid releases) if abnormal
Urinary iodine concentration No established target: track results against WHO’s population adequacy cutoff of 100 µg/L Safety check for iodine deficiency Spot urine reflects recent intake; repeat samples improve individual estimates; cutoff from WHO, whose member governments earn no revenue from it (WHO, 2013)
Serum potassium 3.5–5.0 mmol/L (standard reference range) Safety check for hyperkalemia Only for CKD or potassium-raising drugs; avoid fist clenching, which falsely raises results
eGFR No established target: track change from own baseline Safety check for protein load Creatinine-based; below 60 mL/min/1.73 m² defines CKD stage G3 per KDIGO (KDIGO, 2024); high meat intake can transiently raise creatinine, so pair with cystatin C (an alternative kidney marker) if declining
Bone mineral density (DXA) No established target: track change from own baseline Safety check for calcium shortfall DXA: dual-energy X-ray absorptiometry, a low-dose bone scan; for postmenopausal women and older men; repeat at 2 years on the same scanner
INR No established target: track stability against own prior values Safety check for vitamin K interaction Only for vitamin K antagonist users; check weekly for the first month after switching

Qualitative markers:

  • Satiety between meals and absence of cravings
  • Energy levels across the day and during training
  • Digestive comfort and stool regularity after removing and reintroducing foods
  • Training performance, particularly in high-intensity sessions
  • Sustainability: ability to follow the pattern socially and while traveling

Emerging Research

  • Modified paleo vs ketogenic diet in MS: NCT05007483 randomizes 162 people with relapsing-remitting MS (flare-ups separated by recovery) to modified paleo, time-restricted ketogenic (very-low-carbohydrate) or usual diets; primary outcome: quality of life to 24 months; active, not recruiting, completion December 2026. Positive results would raise the MS item; null results would lower it.
  • Online autoimmune diet course: NCT05057676 tests the investigator’s commercial course on modified paleo, vegetarian or Mediterranean diets and lifestyle; in 100 people with MS it reduced fatigue versus waitlist (Titcomb et al., 2023). Fibromyalgia (chronic widespread pain) and long COVID cohorts end December 2027; positive results would extend it beyond MS, null results would not.
  • Mediterranean vs paleo in rheumatoid arthritis: NCT07438652, a 75-person 12-week trial using a disease-activity score (DAS28) as primary outcome; its registered completion date (July 2026) has passed, the record still lists recruiting, and no results are posted.
  • Lifestyle trial in early MS: NCT04009005, a non-randomized trial in 44 newly diagnosed people, found a modified paleo diet plus breathing exercises and walking not worse than usual care for fatigue and quality of life at 12 months (Saxby et al., 2024); positive 2027 follow-up would strengthen the MS item, null follow-up would weaken it.
  • Weight-stable mechanism trials: Matching paleo and diabetes diets for weight and macronutrients erased the glycemic difference (Fontes-Villalba et al., 2025); larger replications could settle whether food exclusions matter independently.
  • Gut microbiota and TMAO: Cross-sectional data linking long-term strict adherence to higher TMAO (Genoni et al., 2020) need longer randomized follow-up with resistant-starch additions, which could weaken or reverse this concern.
  • Hard outcomes and adherence scores: Cohort meta-analysis suggests lower mortality (Bahrami et al., 2026), yet paleo scores tracked no cardiometabolic changes in a 2-year trial (Rydhög et al., 2025); long trials with events and bone outcomes remain missing.

Conclusion

The paleo diet replaces grains, beans and lentils, dairy foods, added sugar, salt and processed foods with meat, fish, eggs, vegetables, fruit and nuts. For adults willing to follow a demanding eating plan, the evidence shows a dependable short-term pattern: more weight and waist loss than standard healthy-eating advice, lower blood fats and somewhat lower blood pressure, with the clearest gains in people carrying excess weight or early blood-sugar or heart problems. Effects on blood sugar appear to flow mostly from eating less rather than from the specific foods removed, and advantages tend to shrink over two years as people drift from the plan.

The risks are concrete but manageable. Strict versions can leave people short of iodine and calcium, men who do not strength-train may lose muscle along with fat, and menus built heavily on red meat carry the cancer concerns attached to that food. People with reduced kidney function, those who are pregnant, and people taking certain heart, mood or clot-preventing medications face particular cautions.

The evidence base rests on many small, short studies from a handful of research groups, and links with longer life come from studies of everyday eating habits rather than from people following the diet. The diet’s best-known popularizers sell branded books and products, and much of the multiple sclerosis evidence comes from a group whose leader sells courses on her version. Overall, the diet performs about as well as other whole-food patterns, and its value depends on careful planning and long-term adherence.

Top - Benefits - Risks - Protocol