To date there are 16 paleo diet studies (with over 20 published papers), 14 on the classic paleo, and 2 using the stricter auto-immune paleo (AIP) diet. Here is a link to blog posts on the AIP for multiple sclerosis and inflammatory bowel disease.
The Paleo, short for Paleolithic diet continues to be controversial amongst health professionals. It is often touted in the media as being high in animal protein, high fat (especially saturated fat), low carbohydrate and unhealthy because it cuts out entire food groups (grains, dairy, legumes).
Posts such as this one which claim the paleo diet is high in red meat and saturated fat, cutting grains is dangerous, presumably because fibre (gut health) and B vitamins.
I’ve taken a close look at all published paleo diet intervention studies. In this post I focus on studies where the food changes are clearly documented. That is total energy, carbohydrates, fats and protein changes were documented and micronutrients and fat types analysed.
The first thing that stands out on reading these papers is the diet used in the studies is quite different from the perception of what people eat on a paleo diet, and the one in the media.
All these studies use a classic paleolithic diet as recommended by Loren Cordain. Here is an outline of the actual paleo diet used in studies. It is neither very low carbohydrate or high fat:
Foods excluded in all studies, i.e. not allowed on a paleo diet:
- Cereal grains, including corn
- Dairy products
- Legumes, includng soy and peanuts
- Refined fats
- Refined sugars
- Added salt
- Beer
- Soft drinks
- Processed meat – some allowed limited cured meats
- Processed foods in general
Foods allowed, unless stated there is no limit on the amount eaten:
- Lean meat and poultry
- Fish and seafood
- Eggs (often restricted to 2 per day)
- Fruit and berries
- All vegetables including root vegetables
- Some studies allowed one to 2 potatoes a day, others none
- Nuts (40 – 60 grams a day) almonds and walnuts were often recommended
- Avocado
- Dried fruit in small amounts in some studies (1 tablespoon day to 130 grams a day)
- Added oil: olive or rapeseed (canola) oil, 15 grams or 1 tablespoon a day
- Small amounts of honey and vinegar for flavour
- Red wine, from 1 glass a week to maximum of 2 glasses a day
- Unsweetened almond milk
- Salt free spices
- Coffee and tea
Protein, carbohydrate and fat ratios by calories were controlled in some studies, and not others. They were controlled by providing meals, or meal plans. Those that had controlled macronutrient ratios tended to use this ratio:
Protein 30%, Carbohydrates 30%, Fat 40% by calories
Some studies supplied food, others did not. Even those that supplied food allowed participants to eat ad libitum – to appetite, they could add allowed foods if they were hungry.
How did the diets change from baseline or compared to the reference diet (non-paleo diet)?
- Protein stayed the same or increased a little (around 15 grams a day)
- Saturated fat decreased significantly and monounsaturated and polyunsaturated fat increased or stayed the same.
- Total fat did not change much, in some studies it decreased.
- Carbohydrates decreased by about half in most studies, however the lowest carbohydrate intake in the studies was 77 grams per day. The range was 77 to 164g/day. All bar one group were above 100 grams a day.
- Fibre intake tended not to change from baseline
- Food volume and weight stayed the same, even though energy intake reduced.
- Calories decreased by about 25 – 30% (remember people are eating to satiety and not hungry)
- Sodium decreased significantly and potassium increased in all paleo groups
- Micronutrients: Calcium typically reduced, whereas many others increased including B6 and B12
Overall results worth a mention
Compared to baseline or other healthy diets, the paleo diet achieved better results in every singly study in all the following measures
- Satiety and satiety hormones
- Greater reduction in caloric intake without counting calories
- Greater reduction in fat mass
- Greater reduction in waist circumference
- Greater improvements in insulin sensitivity and other markers of pre-diabetes and type 2 diabetes
- Greater reduction in blood pressure
- Greater improvements in blood lipids
Please note and I think it is worth mentioning – these results can only be applied to the type of paleo diet in the guidelines at the top, it does not apply to any other version of the paleo diet or a ketogenic, or other type of low carbohydrate diet.
The Studies
Study 1: Otten, J et al, 2017
Benefits of a Paleolithic diet with and without supervised exercise on fat mass, insulin sensitivity, and glycemic control: a randomized controlled trial in individuals with type 2 diabetes (PDF), Pubmed
Study type and intervention: Randomised to two groups: Paleo diet plus standard care exercise (PD, n=15) and PD plus supervised high intensity exercise, 3x week (PDEX n=14). Ad libitum diet.
Participants: 32 people with overweight with type 2 diabetes, BMI 25 – 40 kg/m2, 29 completed study. Plus an additional reference group of 9 with no intervention
Study length: 12 weeks
Results: Fat mass decreased in both groups: 5.7kg in PD group and 6.7kg in PDEX group. Waist circumference PD -7cm, PDEX -8cm, and insulin sensitivity increased. Blood pressure decreased in both intervention groups without any group difference: systolic,13% in PD and 8% in PDEX; diastolic,10% in PD and 12% in PDEX
How did the diet change compare to baseline?
- Total energy reduced: PD 2064 to 1446 kcal, PDEX 1728 to 1307
- Protein slight increase PD: 83 to 96, PDEX: 77 to 79
- Fat, total PD: 88 to 71g/day, PDEX: 67 to 61 g/day
- MUFA and PUFA increased a little. Total SFA decreased: PD 34 to 15g per day, PDEX 27 to 14
- Total carbohydrates decreased: PD 200 to 127, PDEX: 169 to 77g per day
Study 2: M Ryberg et al., 2013
A Palaeolithic-type diet causes strong tissue-specific effects on ectopic fat deposition in obese postmenopausal women. Pubmed (PDF)
Study type and intervention: Paleo diet ad libitum, pilot study. Women were their own controls, variables at baseline and end of the study were compared for changes.
Participants: 10 post-menopausal, non-smoking women, healthy, overweight with BMI > 27
Study length: 5 weeks
Results: Body weight reduced by 4.6kg; 86.4 to 81.8kg, Hepatic lipid content decreased by 49%. Blood pressure, waist circumference and leptin also reduced.
How did the diet change compared to baseline?
- Energy intake decreased by 25% (2408 to 1888kcal)
- Protein increased by 27%, 105 to 122g per day
- Carbohydrates decreased by 58%, 281 to 118 g per day
- Fats: total 33 to 43.5 g per day. SFA decreased by 57%, 39.2 to 16.7 g per day. PUFA Increased by 122%, 11.6 to 25.8g per day. MUFA 32.8 to 44.9g per day
- Fibre 25 to 26.9g per day
Study 3: C Melberg et al., 2014
Long-term effects of a Palaeolithic-type diet in obese postmenopausal women, a 2-year randomized trial. pubmed
Type of study and intervention: Randomly assigned to paleo Nordic Nutrition Recommendations (NNR) or paleo diet (PD); 35 in each group.
Participants: 70 women (61 analysed) – post menopause, non-smoking, obese (BMI >27), mean age 60yrs, (range 49 – 71) mean BMI 33. at 24 months 27 in PD, 22 in NNR completed
Study length: 24 months, with measures at 6, 12, 18 and 24 months
Results: Both the diet groups decreased their total fat mass: -6.5 and -2.6kg at 6 months; and -4.6 and -2.9kg at 24 months for the PD and NNR group, respectively.
How did the diet change compared to baseline, 6 months and 24 months?
- Total energy (kcal) PD: 2000, 1625, 1599. NNR: 2019, 1660, 1768 (The PD group had a 19% and 20% lower reported energy intake and the NNR group 18% and 12% lower reported energy intake at 6 and 24 months, respectively.)
- Protein PD: 84.4, 93.7, 84.8, NNR: 85.2, 76.5, 76.4
- Carbohydrate: PD; 224, 120, 137. NNR; 222, 181, 190
- Fat: Decreased in NNR by 18g per day and 3.5g per day in PD at 6 months
Study 4: T Jönsson et al., 2009.
Beneficial effects of a Paleolithic diet on cardiovascular risk factors in type 2 diabetes, a randomized cross-over pilot study. (PDF) Pubmed full-text link
Study type and length: Randomised open label crossover, 3 months on Paleo diet (PD) and 3 months on current dietary guideline for diabetes (DD)
Participants: 13 total, 7 started PD first, 6 started with DD. Type 2 diabetes duration average 9 years
Study length: 6 months (3 months each diet)
Results: Paleo diet resulted in lower HbA1c, triglycerides, BP, Weight, Waist circumference, higher HDL, compared to Diabetes diet
Diet differences between PD and DD:
- Paleo diet lower in total energy, 1456 kcal vs 1581 kcal, although it weighed the same. Paleo diet had lower energy density. Mj per gram – PD; 4.7, DD; 5.6
- Lower in carbohydrate, 125 vs 196g
- Protein similar: PD 94 vs DD 90g
- Fats: total PD 68g/day, DD 72g/day. PD higher in UFA. SFA, lower in PD; 19, vs DD 27g
- Nutrients: PD slightly higher in many vitamins, except calcium; lower on PD 356, vs DD 698.
- Glycemic Load: PD meals had a lower glycemic load. 63 vs 111
Study 5: Lindeberg, 2007.
A Paleolithic diet is more satiating per calorie than a Mediterranean-like diet in individuals with ischemic heart Disease. (full) PubMed
Type of study and length: 12 weeks, Randomised Controlled dietary intervention. Mediterranean like diet (MD) compared to paleo diet (PD)
Participants: 29 male with ischemic heart disease (IHD) and with impaired glucose tolerance or type 2 diabetes. Increased OGTT. Waist >94cm. PD n=14, MD n=15
Results: PD, 20% decrease in the oral glucose tolerance test (OGTT) during the first 6 weeks and an 8% decrease between weeks 6 and 12. In the MD group, a 10% decrease of OGTT after the first 6 weeks with no further change at 12 week. After 12 weeks, all 14 subjects in the Palaeolithic group had normal 2 hour plasma glucose values, compared with 7 of 15 subjects.
Waist circumference decrease: PD 5.6cm, MD 2.9cm, Weight loss PD 5.1kg, MD 3.8kg
Diet difference between groups
- Energy intake PD 25% lower. PD5.6MJ, MD 7.5MJ, i.e. PD consumed less energy, for same satiety.
- Protein intake identical,
- Carb PD 43% decrease. PD 134g, MD 231g.
- Total fat: PD 42g/day, MD 50g/day
- GL 47% lower in PD, correlated strongly with cereal intake. PD 65, MD 122. High fruit and lower cereal intake was associated with larger waist cm loss.
- Satiety Quotient: strong trend for greater satiety quotient for energy in the Paleolithic group (2.7 RS/MJ vs. 1.8 RS/MJ, Paleolithic vs. Mediterranean,
- Similar weight of food: PD 1493 ± 607 g/day vs. 1649 ± 273 g/day
- Lower energy density of food in PD: 4.5 kJ/g vs. 5.4 kJ/g, Paleolithic vs. Mediterranean
- Lower salt intake in PD 4.7g/d vs 7.2g/d
Study 6: A Genoni et al., 2016.
Cardiovascular, Metabolic Effects and Dietary Composition of Ad-Libitum Paleolithic vs. Australian Guide to Healthy Eating Diets. A 4-Week Randomised Trial. (PubMed full Text)
Study design: Randomised, non blinded, Australian guide to healthy eating (AGHE) n=17 compared to PD n=22. Primary aim to compare PD with AGHE health effects and nutritional adequacy.
Participants: 39 healthy overweight women, Age 47 + 13y, BMI 27 + 4,
Study length: 4 weeks
Results: PD significantly greater weight loss; 3.2 vs 1.2kg, decreased waist; 3.4 vs 1.6cm, fat loss, lowered systolic BP; 3.3 vs 0.7 mmHg. Nutrients: PD had significantly lower calcium, iodine, sodium, betacarotene, vitamins C, A & E, and red cell folate. Women on paleo had 76% less desire for sugar foods vs 56% on AGHE
Diet differences between AGHE and paleo diet, changes from baseline to intervention diets.
- Total energy, PD from baseline, drop of 1883Kj, AGHA, 1432 Kj
- Protein: Basline to PD +2g, AGHE, -11g/ day
- Carbohydrate: baseline to PD 191 to 103 g/day, AGHE 188 to 163 g/day
- Dietary fibre similar and no change, approx 25g/day
- Total fat: Baseline to PD 71 to 65g/day, AGHE, 79 to 60
- SFA: PD 27 to 19g/day, AGHE; 31 to 22g/day
Study 7: M Osterdahl et al., 2008.
Effects of a short-term intervention with a paleolithic diet in healthy volunteers. (PDF) PubMed
Study type: Pilot study. Aim: assess the effect of a paleolithic diet in a pilot study on healthy volunteers
Study length: 3 weeks
Participants: Healthy volunteers, 20 – 40 yrs, BMI<30. 10m, 10f started, 5m, 9 f completed
Results: Weight reduced by 2.3kg, Waist by 1.5cm, Systolic BP reduced by 3mmHg
Diet changes from normal diet:
- Total energy: reduced by 36%, 2478 to 1584 kcal/day,
- Protein: increased slightly: 84 to 95 g/day
- Carbohydrates: reduced by 53%, 335 to 158 g/day
- Fibre: no change – range 25 -35g per day
- Fat: redcued 82 to 63g/day. SFA 31 to 15g/day
- Nutrients: Vitamin B6 +76%, Vit C +200%, Vitamin E +69%, Vitamin K +45%, sodium -62%, potassium +45%, calcium -53%,
Study 8. R Pastore et al., 2015.
Paleolithic nutrition improves plasma lipid concentrations of hypercholesterolemic adults to a greater extent than traditional heart-healthy dietary recommendations PubMed
Study type and length: 8 months, 2 phase intervention, own controls. AHA heart-healthy dietary guidelines for 4 months, (phase 1) followed by PD 4 months (phase 2)
Participants: 10 male, 10 female with hypercholesteremia
Results: Baseline to end AHA, then end PD phase. There were greater improvements in the PD phase compared to ADA phase in lipids, Total cholesterol -3% on AHA, -20% PD, LDL, -3% AHA, -36% PD, HDL AHA, no change, +35% PD. Weight loss AHA 3.3kg, PD 10.4kg in men. AHA no change, PD 8.1kg
Diet changes between phases
- Same energy: 8700 kJ perday
- Protein PD 123g compared to 91 g per day
- Carbohydrates PD 164 vs 271g per day
- Fat PD 94g, AHA 68g, SFA 24 vs 21, MUFA 44 vs 26.5, PUFA, 19vs 14.6, EPA+DHA 1580 vs 570mg.
- Fibre PD 34, AHA 28
- Sodium PD 2194 vs 2121, potassium PD 5859 vs 3932
This last study measured satiety hormones after single meals
Study 9: F Bligh et al., 2015.
Plant-rich mixed meals based on Palaeolithic diet principles have a dramatic impact on incretin, peptide YY and satiety response, but show little effect on glucose and insulin homeostasis: an acute-effects randomised study. (Full study link) PubMed
Study type: Three different meals were tested in 24 healthy male subjects, each subject tested each meal, randomised crossover. Meals: paleo diet meal 1(PD1) paleo principles modern foods with PCF ratios based on hunter-gatherer studies. PD2 same ingredients, but lower protein to match reference (ref)meal based on WHO guidelines. All 3 meals had 50g available carbohydrates. Carbohydrates in WHO meal was from rice, fruit, vegetables, in PD meals from vegetables, fruit, and mushrooms.
Results: Both PD1 and PD2 had a significant increase in GLP-1, and Peptide YY and improved satiety. The incretin hormone GIP was dramatically lower and showed a very different response curve for both PAL1 and PAL2 in comparison with the REF meal
