Public health recommendations continue to emphasize the importance of overall dietary patterns that support nutrition adequacy and reduce the risk of chronic disease.1 Between 2015 and 2018, approximately 17% of adults reported following a special diet, ranging from weight loss diets to low-carbohydrate diets or diets for diabetes management.2 The popularity of fad diets changes as new diets3 emerge in the public domain. Currently, a small segment (<4%) of the population reports following a low-carbohydrate (Paleo-type diet) or Paleo diet.2 The Paleo diet focuses on naturally raised meat and fish, vegetables, and fruits while avoiding dairy products and grains.4 It also aims to minimize or exclude processed foods but typically does not avoid processed fat and oils.4 Although interest and research on the Paleo diet has increased in the past decade,5 few well-controlled randomized trials exist, and the effects of long-term adherence to a Paleo dietary pattern are unknown.4 Our corresponding work published in the last edition6 details the history of the Paleo diet, noting the inconsistent definitions of the modern-day Paleo diet and the large variation in macronutrient and dietary fiber intake across studies. Thus, evaluating the potential health effects of the diet or making recommendations regarding the healthfulness of this diet is challenging.7 Since the first contemporary Paleo diet book was published in 2002 by Dr Cordain, The Paleo Diet: Lose Weight and Get Healthy by Eating the Food You Were Designed to Eat,8 a variety of recipes and meal planning tools have become widely accessible on the web for individuals who choose to follow a modern-day Paleo dietary pattern.9-11 However, it is unknown how these promoted recipes and meal plans translate into usual nutrient intakes or the nutritional adequacy of adherence to a Paleo dietary pattern. To our knowledge, the nutrient composition of recipes and meal plans from popular Paleo books or online resources and the resulting theoretical nutrient intakes have not been assessed.
Our objective was to assess the relative dietary quality of theoretical, modern-day Paleo meal plans in comparison to US nutrition guidance. In this analysis, we compared food and nutrient levels of theoretical intakes to the Dietary Reference Intakes (DRIs) and MyPlate meal plans based on 30 days of suggested meal plan data.
METHODS
Sources of Theoretical Paleo Diet Information
Respondents in the Adhering to Dietary Approaches for Personal Taste Feasibility Survey reported their sources of diet and nutrition information for their respective diets.12 The Adhering to Dietary Approaches for Personal Taste Feasibility Survey was a brief online survey targeted at self-identified popular diet followers. The survey was conducted over 8 weeks, from July 14 to September 14, 2015. A total of 13 787 participants consented to enroll. After identifying their usual diet from a multiple-choice question, a follow-up, free-text question was presented that asked participants to report their sources of cooking and nutrition information (books and websites) for their diet, as previously described.13 Study participants were able to report 1 or more sources. In response to this question, 925 participants (16%) who had identified as following a Paleo diet provided information on sources of dietary information. We restricted our analysis to a random sample of 200 self-identified followers of the Paleo diet.
Creation of Theoretical Meal Plans
Food and nutrient intakes for the theoretical Paleo meals were assessed using the following steps. First, free-text survey responses were coded for unique responses of recipes and nutrition information (ie, books or websites authored by different Paleo proponents or organizations) among self-identified followers of the Paleo diet. Most responses included at least 2, but often more, unique sources. Next, representative meal plans were drawn from the top 6 diet information sources (top 3 websites and top 3 books most consistently cited) that were reported by at least 5% of the selected respondents. Lastly, a 5-day meal plan from each of the 6 sources was identified or created using provided recipes and instructions by 1 researcher (M.C.K.), for a total of 30 representative days. Our research group previously applied the same methodological approach for developing theoretical vegan/whole-food plant-based diets and discussed the strengths of this approach in an earlier publication.13 The methodological considerations also apply to deriving the theoretical Paleo estimates here. Meal plan selections and quantities were checked by a second researcher.
For comparisons, US Department of Agriculture-compliant meal plans were generated from 21 days of MyPlate meal plans, which have been constructed to fully meet Dietary Guideline recommendations for food and nutrient composition (Sample Two-Week Menus and Sample Menus for a 2000 Calorie Food Pattern).14,15 MyPlate and Paleo meal plan data were collected and analyzed by a single researcher (M.C.K.) using the Nutrition Data System for Research (NDSR) software, version 2016, developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis.16 Meal plans for the Paleo diet were also standardized to 2000 kcal/d to match MyPlate meal plans.
Recipes were entered exactly as written to the greatest extent possible (M.C.K.). Ingredients in the NDSR database were selected to maximize consistency with recipe instructions from the books and websites. To select generic ingredient choices and standard portion sizes when instructions were unclear, NDSR data entry rules were used. For example, when entering meats, the nutrient database offers the option to select "unknown" with respect to whether visible fat is eaten, and this option was utilized for many recipes lacking detailed information. With respect to assigning nutrient values to unknowns, the database defaults to values for the form of the food that is believed to be most commonly consumed in the United States. Contents of meal plans were entered in full and divided by the number of servings to produce single-serving portions. Neither Paleo nor MyPlate meal plans included alcohol. The accuracy of recipes and meal plans entered into NDSR was confirmed by a second reviewer (A.M.).
Dietary supplements were included in the meal plans if they were suggested by at least 2 of the meal plan sources. Applying these criteria, we included magnesium and vitamin D from those sources that recommended them. Three sources recommended magnesium supplements (400 mg/d, 400-600 mg/d, and unspecified generic dose per day, respectively), and 2 sources recommended vitamin D (5000 and 2000-5000 IU/d, respectively). Supplement data were entered using the 24-hour supplement intake module in NDSR. Nutrient levels were calculated as total (food plus supplements) and diet only.
Food and nutrient data from the meal plans were exported from NDSR for descriptive analysis into SAS 9.4 (SAS Institute, Cary, North Carolina). Before calculating the mean theoretical food and nutrient levels, the information from the sources were weighted, using an approach previously described.12 Although this weighted approach was applied to create a more representative estimate of food and nutrient intake, unweighted estimates were similar; hence, the unweighted data (e.g., the average of 30 days) are not presented. The Healthy Eating Index 2015 (HEI-2015),17 an index of overall dietary quality that measures adherence to the 2015 Dietary Guidelines for Americans (DGA),1 was used to assess dietary quality.
We compared the mean nutrient and food group levels from theoretical Paleo meal plans with those from MyPlate, as previously described.13 In addition, the nutrient content of theoretical Paleo meal plans was compared with the relevant DRIs from the National Academy of Medicine18 for nonpregnant female and male adults (ages 19-70 years) for those nutrients labeled as public health concern by the 2020 US Dietary Guidelines (calcium, potassium, dietary fiber, and vitamin D), as well as other nutrients frequently underconsumed or overconsumed.19 When there were differences in recommended intakes across ages 19 to 70 years, the highest value for the DRI was selected for comparison purposes. To determine the nutritional profile of Paleo meal plans, we applied several criteria20 to estimated theoretical nutrient levels as shown in Figures 3 to 5: (1) Estimated Average Requirement (EAR) and Recommended Dietary Allowance (RDA) for vitamin A (retinol activity equivalents, [mu]g), vitamin D (calciferol, [mu]g), vitamin E (mg), folate ([mu]g), vitamin C (mg), calcium (mg), magnesium (mg), and iron (mg); (2) Adequate Intake for potassium (mg) and fiber (g); (3) Chronic Disease Risk Reduction Intake (CDRR) for sodium (g); and (4) recommendations from the 2020 US DGA to limit saturated fat and added sugar to less than 10% of calories. The percentage differences for Paleo meal plans as compared with MyPlate meal plans or DRIs were calculated as (Paleo value/reference value) x 100 - 100.
RESULTS
A comparison of food groups between the theoretical Paleo and MyPlate plans is presented in Figure 1. MyPlate meal plans contain a variety of servings of all food groups. Although the Paleo diet as popularly promoted typically eliminates dairy, one of the major sources identified did include some dairy ingredients in the recipes, thus explaining the 0.3 cup mean intake of dairy. There are notable differences in food groups in comparison to MyPlate meal plans. Theoretical Paleo meal plans provide more total vegetables (5.3 vs 2.5 cup equivalent [eq] of Paleo vs MyPlate, respectively), green leafy vegetables (3.0 vs 0.8 cup eq), red meat (5.3 vs 1.2 oz eq), eggs (2.3 vs 0.7 oz eq), and seafood (3.2 vs 1.2 oz eq), while also supplying fewer legumes (0 vs 0.5 cup eq), whole grains (0 vs 3.4 oz eq), and refined grains (0 vs 3.4 oz eq) (Figure 1). MyPlate recipes specified lean cuts of meat, whereas a number of the Paleo recipes did not specify whether the cut of meat was lean. Variation in individual sources of meal plans produced overlapping 95% confidence intervals for all food groups except dairy, whole grains, refined grains, and legumes.
The mean nutrient contents of MyPlate and Paleo meal plans are presented in Table 1. With respect to macronutrients, mean percentage energy from carbohydrate, protein, and fat was 54:19:29 versus 18:24:60 for MyPlate and Paleo, respectively. Total grams of carbohydrate for Paleo meal plans were 91, which is below the RDA of 130.20 Percentage energy of added sugar was 6 for MyPlate and 1 for Paleo. Total dietary fiber was 25 g for Paleo compared with 28 g for My Plate (Table 1). Percentage energy from total protein, plant protein, and animal protein was 19%, 7%, and 12% kcal for MyPlate and 24%, 4%, and 20% for Paleo, respectively (Figure 2Table 1). Saturated fat content was 8% for MyPlate and 19% for Paleo (Figure 2Table 1). Percentage of saturated and polyunsaturated fats in the Paleo diet contributed 19% and 10%, respectively. The ratio of unsaturated to saturated fat was 3:1 for MyPlate compared with 2:1 for theoretical Paleo diets. The HEI-2015 score for estimated theoretical Paleo diets was 72 of 100 available total points, which represents the points that align with the MyPlate plans (MyPlate plans were designed to achieve a score of 100). Median food and nutrient data was similar to mean data for both MyPlate and Paleo meal plans (Table 2).
Comparisons of Paleo meal plans with DRIs are shown in Figure 3 (RDAs and EARs for men), Figure 4 (RDAs and EARs for women), and Figure 5 (AIs and CDRRs). Estimated nutrient levels, including vitamin D and magnesium supplements, from the theoretical Paleo meal plans were as follows: meal plans meet or exceed the RDAs for vitamin A, vitamin D, vitamin E, folate, vitamin C, magnesium, and iron for men, but not for iron among women ages 19 to 50 years or calcium for men or women ages 51 to 70 years. Paleo meal plans do not meet the EARs for calcium or the AIs for potassium or fiber. Estimated levels of sodium are higher than the CDRR in both men and women ages 19 to 70 years. Levels of saturated fat exceed the recommendation, whereas added sugar levels fall within the recommendation (Figure 5).
DISCUSSION
The purpose of this analysis was to estimate the theoretical food and nutrient composition of common Paleo recipes and meal plans and to compare a theoretical Paleo diet with current recommendations. Although there is broad variation and a wide variety of definitions of common Paleo diets,4 we applied a unique approach to construct meal plans from recipes derived from nutrition and cooking sources being used by self-identified Paleo diet followers in a large online survey.
Based on the results of this analysis, Paleo diets have both advantages and drawbacks with respect to dietary quality as compared with US consumption patterns. National dietary intake data place the average American adult at a score of 59 on the HEI-2015,17 in comparison with the MyPlate target of 100. The HEI score for the theoretical Paleo meal plans was 72. Paleo meal plans also exceed the levels in the MyPlate meal plans for total vegetables, green vegetables, and nuts and seeds. According to data from the National Health and Nutrition Examination Survey 2009-2010 survey, roughly 14% of total energy consumed in the US is from added sugars,21 and only 30% of calories comes from unprocessed or minimally processed foods including meat or dairy, grains, legumes, and fruits and vegetables.22 The theoretical Paleo plans are composed almost entirely of unrefined foods and thus achieve levels of added sugar well below the Dietary Guidelines threshold of 10% for total energy23 and American Heart Association thresholds of 100 kcal/d for women and 150 kcal/d for men.24 In addition, the micronutrient profile of Paleo meal plans would make it relatively easy to achieve the EARs and RDAs for most nutrients, including vitamins A, C, E, folate and, when including supplements, magnesium and vitamin D, all typically underconsumed nutrients.19
Other aspects of the Paleo meal plans are cause for concern in terms of cardiovascular disease risk. The meal plans have a typical "low-carbohydrate" profile, limiting whole grains and legumes and emphasizing meat, which translates into a theoretical estimated 18% of calories from carbohydrate, 60% total fat, and 19% saturated fat, with lower fiber levels. Both carbohydrate and fat levels fall well outside of the acceptable macronutrient distribution range, which is 20% to 35% for fat and 45% to 65% for carbohydrate.18 The RDA of 130 g of carbohydrate is not met with the Paleo plans containing only 91 g of carbohydrate. The theoretical fiber content is 25 g/2000 kcal, which, although higher than current estimates of fiber intake in the United States, is somewhat under the Adequate Intake of 14 g/1000 kcal.20,25 This is also relevant for heart health, as greater intakes of both fiber26 and fiber-rich foods27 have been associated with lower cardiovascular mortality risk. The theoretical fiber content is, overall, consistent with previous reported intakes derived from Paleo intervention studies (21 g,28 25 g,29 and 32 g30). With such a high proportion of total calories coming from fat, there is limited room to add carbohydrates within an approximately 2000 kcal/d meal plan, and, if strictly adhered to, such intakes could induce ketosis in some individuals.31 There is some interest in low-carbohydrate and ketogenic diets for weight loss32 and/or blood glucose control33 in the short-term. However, there may be a variety of potential health consequences31,34 associated with long-term adherence to a very-low-carbohydrate diet.31,35
Estimates of saturated fat intake are higher than recommended limits from the current Dietary Guidelines for Americans recommendations (<10% kcal/d from saturated fat)23 and American Heart Association guidelines, which have an even lower threshold for high-risk individuals (5%-6% saturated fat), coupled with a low unsaturated-saturated fatty acid ratio.36 In terms of food composition, both total meat and red meat intakes in the Paleo meal plans are in excess of accepted guidance to reduce cardiovascular risk.36
It is possible that in real-world settings, the intake of saturated fat may be lower than our estimate of 19% of energy. In some intervention studies, participants assigned to a Paleo diet as part of a dietary intervention consumed saturated fat ranging from 12% of energy intake after a 4-week trial,39 to 11% in a 3-month crossover study (n = 13),28 and to 6% in a 3-week trial (n = 6).30 However, although reported saturated fat intakes in these intervention studies are lower than our meal plans' estimates, saturated fat intake in these promoted Paleo meal plans is still higher than current recommendations to promote heart health.1,24,37
Drawbacks in terms of micronutrients include inadequate calcium and potassium levels as well as sodium levels in excess of recommendations.19 Because of the exclusion of dairy, Paleo diets rely on other sources of calcium, including green leafy vegetables and fish with edible bones; however, the meal plans fail to meet the RDAs and EARs for calcium for both men and women. This is consistent with the very low levels of calcium previously reported in intervention trials using the Paleo diet (ranging from approximately 350 mg28,29 to 400 mg30). The EARs for vitamin D were achieved through emphasis on fatty fish, and the RDAs were achieved through inclusion of vitamin D supplements. Lower potassium intake and higher sodium intake have been associated with elevated risk of hypertension, and the sodium-potassium ratio may also play a role.38-40 Adherent followers of Paleo diets should be advised by their healthcare practitioners that they may be at risk of calcium and potassium deficiencies without supplementation. None of the meal plan sources in our analysis recommended calcium or potassium supplements. This analysis suggests that the Paleo diet, followed closely over the long-term, might lead to problems with bone health, particularly among older Americans who are at greater risk for osteoporotic fractures.41
Our analysis had several strengths, which include using 30 days of meal plans; utilizing sources of recipe information cited by free-living Paleo followers; and taking a random sample (n = 200) from among a relatively large total sample (N = 925). However, this methodology of using derived meal plans is also a limitation in that our analysis cannot speak to actual intakes of Paleo followers, and the nutrient composition of Paleo diets in intervention settings and free-living populations may vary from targets estimated here. The validity of the sources of nutrition and cooking information is also limited to those in our self-selected sample (largely White women with Internet access) and who may or may not represent typical Paleo followers. Also, comparisons between MyPlate meal plans and the Paleo meal plans represent theoretical comparisons; it is important to note that MyPlate meal plans were constructed specifically to meet DGA recommendations. Finally, it is possible that theoretical estimates of fat and saturated fat may be overestimated because of the fact that lean meat was not specified in most Paleo recipes.
This work helps to characterize the modern-day Paleo diet followed by free-living individuals and was a first step toward comparing population targets with actual intakes. Assessing the degree of dietary adherence to any diet requires reference data, and as such, these data can serve as a reference point for clinicians or practitioners to estimate patient or client adherence to the prescribed diet. The data generated from these theoretical diets may help health professionals, particularly those who practice nutrition education, to better understand the nutritional benefits (high intake of fruits and vegetables, less added sugar) and shortcomings (higher total and saturated fat, inadequate calcium and potassium) of the Paleo diet.42-44 It may be more helpful to recognize these specific advantages and incorporate only select components of the diet as opposed to embracing the principles of the Paleo diet in totality, which would include high levels of meat and saturated fat consumption.
CONCLUSION
As Paleo diets are of current interest among the public, there is a need to assess dietary quality and potential health implications of promoted diet guidance. Meal plans derived from popular sources of Paleo recipes offer some improvement over the typical American diet with respect to the consumption of refined grains, added sugars, and fruit and vegetable consumption. However, even with these advantages, popular Paleo diet meal plans result in high saturated fat levels, well above recommendations, which is a concern for cardiovascular risk, and the carbohydrate, fiber, calcium, and potassium levels do not meet DRIs. Overall, adoption of the Paleo diet should be approached with caution. Future research should capture actual intakes and assess adherence by comparing targeted versus actual intakes.
REFERENCES