The DAA recently did a three part series on the Paleo diet. Here are my thoughts on it
What is Paleo
Part 1 opens up with a description of (Boyd Eaton’s) Paleo followed by some cliché criticisms and implies that Paleo is a re-enactment
· There was not one Paleo diet
· People kept changing and adapting after the Paleo period
· Paleo is expensive and hard to sustain.
That there was not one Paleo diet isn’t really an argument against Paleo, but rather against dogmatic, one-size-fits-all approaches (such as the Australian dietary guidelines). Even though there have been adaptations following agriculture, it’s important to recognise that anatomically modern humans have been hunter-gatherers for ~95% of our time on earth and that not all people have had much adaptation to the foods from agriculture. Both of these points have been acknowledged and discussed in the Paleo community.
They cite this paper as evidence that Paleo is expensive and that a “9.3% increase in income is needed to achieve nutrient targets (except calcium)”. The title of the paper is ‘the feasibility of a Paleolithic diet for low-income consumers’, which tells you they are modelling for on low income consumers. They set a budget of $3.89 per day (US dollars in 2001, equal to $4.91 US dollars in 2010) for food made at home. They found that by choosing low cost foods you would need 9.3% more ($4.25 per day) in the budget to meet micronutrient requirements (except calcium). The authors of the paper then conclude that you need a 9.3% increase in income, but an increase of 9.3% in the food budget wouldn’t translate into an increase of 9.3% in income, unless 100% of earnings were spent on food. So income would only need to increase by a fraction of 9.3%, perhaps 2-3%. It should also go without saying (but doesn’t) that this study was modelled on low-income consumers and does not necessarily apply to the middle-class.
They finish off part 1 with ‘Australians need to eat more fruit and veg’, which I agree with, but what’s the relevance? If anything this is an argument in favour of Paleo. Grains are the centrepiece of the standard Australian diet (cereal for breakfast, sandwiches for lunch and often a pasta/rice dish for dinner). When grains are substantially reduced it creates a void that needs to be filled. Paleo can be thought of as meat and veg with a piece of fruit at 2 or all 3 meals, which would greatly increase fruit and veg intake. Also, in the clinical trials the Paleo group ate more fruit and/or veg  .
The first section of part 2 discusses the pros and cons of Paleo in relation to the food groups. It’s what you would expect: grouping legumes with vegetables, SFA is bad, some evidence from observational studies against fatty meats and in favour of whole grains and the ‘dairy is important for calcium’
The evidence used against SFA in relation to CHD shows no benefit when SFA is replacing with carbohydrate or MUFA only found a benefit when SFA is replaced by PUFA. However, the evidence for this is very poor as the trials that suggest a benefit in replacing SFA with PUFA for CHD were poorly controlled and have confounding variables or were part of a multifactorial intervention (see here). Also, the Cochrane and the Mozaffarian meta-analyses (which are used in favour of replacing SFA with PUFA) didn’t find any benefit in replacing SFA with PUFA for total mortality  .
In the summary they say “…the diet fails to provide all nutrients as per current recommendations and excludes foods and whole food groups”.
Paleo may often not meet the RDI for calcium without dairy, but the RDI for calcium is set in context of sub-optimal vitamin D levels and vitamin K2 intake. Getting more sunlight would increase vitamin D, which would improve calcium absorption and Paleo would often be higher in vitamin K2 (found almost exclusively in animal fats), which is far more effective than calcium at reducing fractures in RCTs  . On the other hand, since the dietary guidelines don’t recommend many eggs and much meat, and don’t emphasise organ meats, a diet based on the dietary guidelines is likely to have insufficient choline.
If the purpose of food groups is to group foods with similar nutrient profiles together so one could get adequate nutrient intake by eating foods from food groups with complementary nutrient profiles, then the current food groups don’t make sense. Mature legumes shouldn’t be grouped with vegetables, grains aren’t unique and can be easily substituted and meat and dairy substitutes poorly match the nutrient profile of the thing they are meant to be (I discussed this a while ago (legumes, dairy, meat)).
Paleo Diet Studies
At the end of part 2 they mention the Paleo diet trials and how they are few and small in number, which is a problem I agree with. They discuss this trial in more detail, saying:
“The authors concluded that the Paleo diet was found to be more difficult to adhere to and each diet was found to be equally satiating, although there were per calorie differences. Weight loss was not recorded in this study.”
This is either incorrect or misleading. The trial the author is talking about had two publications, but the author seemed to have only read the second, as the first publication reports weight and BMI .
Yes, the participants were “equally satiated on both diets” , but since the Paleo diet group were consuming fewer calories  they experienced greater satiety per calorie eaten . This measure of satiety per calorie is extremely relevant to the obesity epidemic as a limiting factor of weight loss is that people get hungry, therefore you would expect foods/diets with greater satiety per calorie to aid weight loss and maintenance. But the DAA ignore that and merely say: “although there were per calorie differences”.
Although more participants found the Paleo diet more difficult to adhere to (P = 0.02), more participants thought the Paleo diet was effective for weight loss (P = 0.02) and almost significantly more participants found the Paleo diet to be more satiating (P = 0.06) and more effective at reducing blood glucose (P = 0.08) . It’s important to not view the perception of ‘difficulty to adhere to’ in isolation. People make decisions based on costs and benefits. The path of least resistance often isn’t the best one and achieving results usually takes some work. The far more relevant question is: do the ends (weight loss, etc) justify the means (adhering to a diet)? That hasn’t been asked in this study but my guess it that the outcome would be a lot less negative than what the DAA depict and probably even positive. Also in my opinion the Paleo diet used in those trials is unnecessarily restrictive (salt, fatty meat, starchy vegetables) and I would like to see the results (weight loss, metabolic markers, etc) and perceptions of adherence in a Paleo trial that doesn’t include unnecessary restrictions, not that I’m holding my breath.
“…it must be recognised that these positive changes were not independent of weight loss”
I don’t understand why this is considered so important. If a given diet promotes weight loss, thereby improving improves metabolic risk factors, because it is more satiating per calorie, then the effect is still causally related to the diet rather than deliberate calorie restriction. In this instance you can say ‘not independent of weight loss’, but should also emphasise how the weight loss occurred (deliberate vs. spontaneous) so as to have a balanced commentary.
“There are problems with the dietary methodology in these studies (i.e. no nuts included in the control Mediterranean diet)…”
I agree that the Paleo diet trials have limitations, but I challenge you to find a diet trial that doesn’t. Anyway, I found it odd that of all the differences between the Paleo and Mediterranean diet (the author is now discussing this trial) they chose nuts. In this trial the Paleo group did eat significantly more nuts (P = 0.02) but it was such an insignificant part of both diets (Paleo 10±12g, Mediterranean 1±3g) . Whereas the main dietary differences in this trial were higher fruit and meat in the Paleo group and higher grains and dairy in the Mediterranean group .
The ‘High’ Protein Australian Diet
There isn’t much to say about part 3, except for this:
“While the Paleo diet focuses on fresh foods, as do the ADGs, there is too much emphasis on protein foods for most people. It is rare that individuals need more than 1g of protein per kilogram of lean body mass and currently most Aussies eat double this recommendation! The focus in Australia should be on getting enough high iron food choices in our diets, because while we all seem to get enough protein, there are some key population groups (mostly girls and young women) who do not get enough iron.”
The RDI is the “usual intake at or above this level has a low probability of inadequacy”  and the RDI for protein is approximately 10% of total calories. Australians eat about 16.5% protein , which is why the standard Australian diet is often described as ‘high protein’. However, in the same document as the RDI there’s also a range of protein intake recommended for reducing chronic disease risk which is between 15-25% of total calories. As the average Australian intake is at the lower end of this range, you could argue that it would be better if we got more protein. I also thought this was an odd comment to make given that: (1) iron is plentiful in most Paleo diets; (2) legumes were promoted as being high in protein; and (3) obesity was discussed at the end and high protein diets aid in weight loss