Tuesday, February 28, 2012

DGA 2011 - Submission

I have made my submission for the draft Australian Dietary Guidelines, you can download it here

It's a 51 page PDF with about 150 references.  It contains several of the posts on the blog such as

As well as some extra stuff.

Sunday, February 19, 2012

Paleo vs DGA

New Food Groups
Australian Dietary Guidelines - Good
Australian Dietary Guidelines - Bad
6 Servings of Vegetables and Legumes/Beans
75g Spinach
75g Tomato
75g Carrot
75g Cos Lettuce
150g Snap Beans (2 Serves)
75g Spinach
75g Tomato
75g Carrot
75g Cos Lettuce
75g Chickpeas
75g Lentils
150g Potato (2 Serves)
150g Corn (2 Serves)
75g Chickpeas
75g Lentils
2 Servings of Fruit
150g Apple
150g Banana
150g Apple
150g Banana
60g Raisins (2 Serves)
6 Servings of Grain (Cereal) Foods, Mostly Wholegrain
150g Potato
150g Sweet Potato

120g Wholegrain Wheat Flour/Wheat Durum (4 Serves)
60g Brown Rice (2 Serves)
90g Wholegrain Wheat Flour/Wheat Durum (3 Serves)
30g White Flour
60g White Rice (2 Serves)
3 Servings of Lean Meat, Poultry, Fish, Eggs, Legumes/Beans and Nuts/Seeds
200g Lamb Chops (2 Serves)
115g Salmon
115g Salmon
75g Kidney Beans
30g Almonds
75g Kidney Beans
60g Almonds (2 Serves)
2.5 Servings of Milk, Yogurt, Cheese and/or Alternatives (Mostly Reduced Fat)
300g 5 Eggs (2.5 Serves)
(+ Vitamin D)

120g Ricotta Cheese
100g Yogurt (0.5 Serves)
250mL Milk
120g Ricotta Cheese
100g Yogurt (0.5 Serves)
250mL Milk

Vegetables and Legumes/Beans: The vegetables are kept constant between the NFG and DGA+.  The point of difference is that DGA+ recommends mature legumes, while NFG doesn’t recommend mature legumes but considers immature legumes to be similar to non-starchy vegetables.  Seeing as corn is considered a starchy vegetable by the guidelines, despite it being a nutritionally inferior grain, I gave it to DGA-, along with mature legumes and potato.

Fruit: Fruit is kept constant between the NFG and DGA+.  125mL of fruit juice is considered a serve of fruit by the guidelines.  I didn’t have data for fruit juice so I used the least nutrient dense fruit as a proxy for the low nutrient density of juice in the DGA- diet.  (Just as dairy is reduced to calcium, fruit is reduced to vitamin C).

Grain (Cereal) Foods, Mostly Wholegrain: The DGA+ has all wholegrains.  NFG places other foods in this category and suggests starchy vegetables are preferable to seed foods due to their lower toxicity.  So NFG has 3 serves of potato and 3 serves of sweet potato.  Each serve is 50g.  DGA- has 3 wholegrain serves and 3 refined grain serves.  The 4 serves of wheat reflects the high consumption of wheat.

Lean Meat, Poultry, Fish, Eggs, Legumes/Beans and Nuts/Seeds: The guidelines suggest legumes, nuts and seeds are alternatives to meat and eggs, so I replaced meat and eggs with seed foods in the DGA diets.

Milk, Yogurt, Cheese and/or Alternatives (Mostly Reduced Fat): NFG suggests alternatives to dairy can be other animal foods as well as vitamin D.  The alternative was eggs.

The differences between the guidelines and NFG are:

Immature Legumes
Mature Legumes
Starchy Vegetables
Mature Legumes, Nuts and Seeds

To compare them, I'm using my nutrient database

At the end of this NFG came to 1,632 calories, DGA+ came to 2,499 calories and DGA- came to 2,550 calories (this is without added fats/oils and extras).  So I adjusted the nutrient intakes of the diets as if there were 2000 calories.  Here are the results for the micronutrients.  The data is the % of the NRV.

Vitamin C
Vitamin B1
Vitamin B2
Vitamin B3
Vitamin B5
Vitamin B6
Vitamin B9
Vitamin B12
Vitamin A
Vitamin E
Vitamin D
Vitamin K1

What’s evident from calcium is that it is very difficult to meet the RDI for calcium without dairy.  As I pointed out in the response to section 2.5, the RDI for calcium should depend on vitamin D blood level of the individual.  An ideal vitamin D blood level of roughly 90 µmol/l decreases the calcium needed to 606 grams, the amount needed may be lower still if the RDI for calcium was set to compensate for widespread low vitamin D in Australia and in the context of a low phytic acid diet, which NFG is.  Seeing as NFG also promotes healthy vitamin D levels, calcium should not be an issue for it.  Among women and sedentary men, who can’t eat 2,000 calories they won’t even be able to meet the RDI for calcium with dairy.

Seed foods are rich in phosphorus as an important building block towards the metal chelator, phytic acid.  80% of the phosphorus in seed foods may be bound up in undigestable phytic acid, so it is not absorbed.  The two DGA diets contain many seed foods and get much of their phosphorus this way.  However if the measurement was which diet has the most bioavailable phosphorus NFG will likely be ahead of the DGA diets.

Even though NFG is richer in sodium it does not matter because salt (sodium chloride) is so easy to come by.  I won’t include sodium when comparing these diets.

It’s odd that DGA+ had slightly more zinc than NFG, even though NFG was richer in animal foods, which are good sources of zinc.  Once again if zinc bioavailability is taken into account NFG will be ahead of DGA+.

“Zinc found in animal products, crustaceans and molluscs is more readily absorbed than zinc found in plant foods. In contrast, legumes and unrefined cereals contain phytates that reduce zinc absorption.” [1]

NFG seems low in manganese, but the opposite could be true – that the NRV is too high.  The NRV for manganese is an AI based on the median intake of the Australian population.  As Australians eat a plant based diet and plants are the main source of manganese, Australians therefore consume a fair bit of it.  To compare this with other recommendations, the USDA AI for manganese is only 2.3 for men and 1.8 mg for women [2].  Both Australian and US recommendations were AI and based on the median population intake.  Seeing as the two country’s diets are similar it might have something to do with the concentration of manganese in soils.  I’m using data from the USDA, so data from Australian food might show higher levels.  The amount of manganese in NFG doesn’t seem to be an issue.

The low intake of choline in both DGA diets is due to our high requirement for choline.  Without eggs, organ meats or a fair bit of muscle meat it’s very difficult to meet this requirement for choline.  Most Australians are stuck in the unscientific dietary cholesterol and saturated fat phobia, meaning the average Australian won’t consume choline rich foods and runs the risk of choline deficiency.  The Australian and USDA AI for choline are the same and are based on what is needed to prevent liver damage rather than median intakes [3].  This is an issue for the dietary guidelines.

Vitamin D is a little out of place as we should be getting most of it from the sun.  I am going to include it because the recommended AI is far too low and fear mongering to keep us out of the sun is responsible for the low vitamin D in Australia.

The worse quality DGA diet (DGA-) was low is a few other nutrients.  Vitamin C was low, but I was too lazy to include orange juice as fruit serves (which was the plan) and that would have made it adequate.  Vitamin B3 (niacin) and B12 was low, which is what you would expect from a diet without meat.  Vitamin A was low because of the lack of carrots, sweet potato, eggs, liver, other vegetables, etc.  A diet of white foods such as grains, potatoes and reduced fat, grain-fed dairy (because vitamin A is fat soluble) may be lacking in some nutrients such as vitamin A.  K1 was low because of the lack of non-starchy vegetables especially the green leaves.  The DGA- diet had no long chain omega 3 because it didn’t include fish and had very little quality animal fats from meat or eggs.

NFG was a better source of 14 out of the 22 nutrients when compared with the DGA diets, which has 8 out of 22.  On average NFG was 37% higher in nutrients than DGA+ and 97% higher in nutrients than DGA-.  I’m now going to adjust for vitamin D and consider calcium to be a non-significant difference, hand over phosphorus and zinc to NFG due to superior bioavailability and consider vitamin E to be a non-significant difference.  Now NFG is a better source of 15 out of the 22 nutrients, 2 are tied and DGA has 5 out of 22.

There are other factors in favour of NFG.  There are fewer problematic dietary lectins, such as those found in grains and legumes, which damage the intestinal lining, increase intestinal permeability and decrease nutrient absorption.  I’m using USDA data for grain-fed animals, which could be improved nutritionally if they were grass-fed.  NFG allows consumption of liver, arguably the most nutrient dense food.

Overall NFG is better than the current recommendations both on paper and even more so when other factors are considered.  That being said current recommendations have an adequate nutrient content except in choline, which can lead to liver damage.  At the very least the guidelines should aim to increase choline intake by recommending eggs, organ meats and more muscle meat.

Sunday, February 12, 2012

DGA 2011 - Meat, Eggs and Alternatives

2.4 Lean meat and poultry, fish, eggs, nuts and seeds, and legumes/beans

Animal-Based vs Plant-Based Nutrition

“The ‘lean meat and alternatives’ food group is diverse, both nutritionally and biologically. The foods in this group have traditionally been seen as ‘protein-rich’, but they also provide a wide variety of other nutrients which may be more important in the typical high-protein Australian diet.”

Just like the section on dairy, this food group has been reduced to being protein-rich.  The so called ‘high protein Australian diet’ is where energy derived from protein comes to about 16.7% of total energy [1], which is towards the lower limit of the optimal protein range (15-25% of total calories) to lower chronic disease risk [2].

Seeing as one of the main nutrients of this group is protein it would be a good idea to make sure these foods are rich in protein and their protein is high quality.  Most nuts contain less than 10% of protein as total calories, which is below the RDI for protein.  Almost all nuts and seeds (except for pumpkin seeds and perhaps a few others) contain less than 15% of protein as total calories, which puts them below the optimal protein range to lower chronic disease risk [3].  With this information nuts and seeds already don’t belong in the category of protein-rich foods and should therefore not be considered a good source of protein (because technically all foods are a source of protein or ‘protein source’).

A common practice in vegetarian circles is to combine lysine poor grains with methionine and cysteine poor legumes.  This results in a less unbalanced meal.

Methionine and cysteine are used to synthesise other compounds that have many health benefits, so replacing meat with legumes will likely decrease the synthesis of these compounds.  Methionine or lysine can be used to synthesise carnitine.  Carnitine lowers triglycerides and increases fat metabolism by transporting fatty acids into the cell [4], decreases oxidised LDL [5] and increases mitochondrial biogenesis when used with lipoic acid [6].  Methionine, arginine and glycine can be used to synthesise creatine.  Creatine is used in the anaerobic creatine phosphate energy system, can improve exercise performance, increase muscle mass [7] and improves cognition [8].  Cysteine, glutamic acid and glycine can be used to synthesise glutathione, sometimes called ‘the master antioxidant of the cell’.  (Cysteine is the rate limiting amino acid in glutathione synthesis [9]).  Glutathione is needed to form glutathione peroxidase and glutathione transferase, enzymes used in reducing free radicals and detoxification.  Cysteine can be used to synthesise taurine, which has a number of functions and health benefits such as reducing insulin resistance [10], triglycerides and ApoB (a marker for LDL and VLDL particles) [11].

Histidine and alanine can be used to synthesise carnosine, which buffers acidity, increases athletic performance [12], blocks protein glycation and chelates heavy metals [13].

Carnitine, carnosine, creatine, glutathione and taurine are synthesised from amino acids, so I will call them collectively the ‘amino acid derivatives’.  In addition to being a good source of the required amino acids, animal foods are rich in the amino acid derivatives as preformed nutrients (carnitine [14]) (carnosine [15]) (creatine [16]) (taurine [17]).  Due to not eating meat, vegetarians and vegans have lower levels of these nutrients in their body (carnitine [18] [19]) (carnosine [15]), (creatine [16]), (taurine [20]).  Glutathione is also found in fruits and vegetables [21] and provided vegetarians consume enough protein (read: methionine and cysteine) they are likely to have sufficient levels [22].

Although data for the concentration of the amino acid derivatives in food is hard to come by, their functions can give us a rough indication.  In addition to all the health benefits above, carnitine (when used with lipoic acid) [23], carnosine [24], creatine [25] and taurine [26] [27] are antioxidants found in mitochondria and also support mitochondrial function.  Mitochondria are responsible for generating most of the ATP.  Energy production tends to be highest in the brain, liver, heart and skeletal muscle, so these tissues are likely to have many mitochondria, therefore the amino acid derivatives are likely to be found richly in foods originating from those body parts.  This is consistent with the limited data available when their other functions are also considered (for example taurine and kidney function).  Plant foods are likely not good sources of these nutrients because plants produce much less energy than animals, hence fewer mitochondria.  The same can be said for other mitochondrial antioxidants such as coenzyme Q10, which also has an essential role in the electron transport chain.  CoQ10 is found mostly in organ meats, muscle meat and some vegetable oils [28].

When considering the value of nutrients, those that are essential and those that the body manufactures are likely to be more important.  The amino acid derivatives are highly valued nutrients as they are roughly half sourced from diet and half from synthesis.  It’s possible to have a deficiency in those kinds of nutrients.  For example, glutathione deficiency is followed by mitochondrial damage [29], which can lead to neurodegenerative disorders such as Parkinson’s disease [30].  Another example is that statin drugs inhibit CoQ10 synthesis, so unless additional CoQ10 is consumed to compensate statins can cause a CoQ10 deficiency, which can lead to mitochondrial dysfunction and type 2 diabetes [31].  On the other hand phytochemicals are neither essential nor manufactured, and there’s no such thing as a polyphenol deficiency.  Instead polyphenols (a large group of phytochemicals) are often antipredation mechanisms and mildly toxic, but we gain some benefits from them through a hormetic response that up regulates our own antioxidant activity to protect against the oxidative potential of polyphenol [32].  The guidelines seem to be biased in favour of plant foods as they discuss phytochemicals thoroughly but neglect to mention the amino acid derivatives and CoQ10.

Protein is widely considered to be healthy, yet our healthy protein intake limited to 15-25% of total calories due to potential ammonia excretion [2].  Seeing as we are limited in our protein intake we should aim to maximise protein quality.  Animal protein is more bioavailable, has optimal amino acid ratios and is a rich source of the amino acid derivatives.  Therefore animal protein should make up as much of the protein calories as possible and plant protein should be considered poor sources of quality protein.

“Important nutrients include iodine, iron, zinc, vitamins, especially B12, and essential fatty acids including long chain omega-3 fatty acids.”

“Nuts and seeds are rich in energy (kilojoules) and nutrients, reflective of their biological role in nourishing plant embryos to develop into plants. In addition to protein and dietary fibre, they contain significant levels of unsaturated fat, although this varies within the category. Nuts are also rich in polyphenols and micronutrients, including folate, several valuable forms of vitamin E, selenium, magnesium and other minerals.”

“Legumes/beans, including lentils, tofu and tempe, provide a valuable and cost efficient source of protein, iron, some essential fatty acids, soluble and insoluble dietary fibre and micronutrients.”

Not only are plant foods poor alternatives to animal food in terms of protein quality, but also in micronutrients.

Nuts and seeds are not rich sources of folate and nuts are not a rich source of selenium [3].  (Brazil nuts are very rich in selenium, 1 nut provides more than the RDI).  Nuts and seeds require vitamin E as it is quickly depleted to protect their high quantities of delicate PUFA from lipid peroxidation [33].  Many minerals in nuts and seeds have reduced bioavailability, but they are still likely to be good sources of many minerals.  Mature legumes are mediocre sources of micronutrients and poor sources of most vitamins.  See the response to section 1.1.

The essential fatty acids in legumes, nuts and seeds are either from linoleic acid or alpha-linolenic acid.  They contain none of the highly sought after long chain omega 3s.  The PUFA in most nuts and seeds is almost exclusively linoleic acid.  Most legumes are also high in linoleic acid and have a high omega 6:3 ratio, including soy, which is about 8:1 [3].  High linoleic acid diets are also associated with insulin resistance and cancer [34] and the resulting high omega 6:3 ratio is pro-inflammatory and associated with poor health outcomes [35].  The alpha-linolenic acid in these foods (especially flaxseeds and walnuts) isn’t very helpful.  It is so poorly converted to DHA that DHA should be considered an essential nutrient.  This is reflective of our ancient diets, rich in wild/grass-fed meat and fish [36].

Legumes, nuts and seeds are not good alternatives to meats because they don’t share many of the same nutrients such as vitamin B12, long chain omega-3 fatty acids and the amino acid derivatives (previously discussed) [3].  There are more nutrients that animal foods are rich in, but most plant foods aren’t.  These include zinc, selenium, riboflavin, niacin, vitamin B5 [3], vitamin B6 (due to poor absorption of plant forms [37]), biotin [38], choline [3], vitamin K2 [39], conjugated linoleic acids (found only in food from ruminants and perhaps eggs) [3], coenzyme q10 [28].  Plant foods are also rich in several nutrients that most animal foods tend to be low in.  Below is a table to show the nutrients animal foods tend to be rich that plant foods tend not to be rich in and vice versa.

Vitamins and Minerals
Animal but not Plant Foods
Plant but not Animal Foods
Vitamin B5
Vitamin C
Vitamin B6
Vitamin E
Vitamin B12
Vitamin K1
Vitamin K2

Other Nutrients
Animal but not Plant Foods
Plant but not Animal Foods
Amino Acid Derivatives
Phenolic Compounds
Long Chain Omega 3 Fatty Acids
Conjugated Linoleic Acids
Lauric Acid
Coenzyme Q10
Soluble Fibre

As omnivores, to be well nourished we need a diverse range of nutrients, sourced from both animal and plant foods.  It is common that animal foods are rich in a nutrient that plant foods are low in and vice versa.  Therefore the only alternatives to an animal food should be other animal foods, and likewise the only alternatives to a plant food should be other plant foods.

Animal Food Consumption in Studies – Definitions and Confounding

Despite meats being good sources of a number of animal-source nutrients, the dietary guidelines cites research suggesting meats are associated with an increased risk of some diseases.  What’s interesting is that meats are positively associated with cardiovascular disease, but dairy is negatively associated cardiovascular disease.  One of the proposed mechanisms by which dairy reduces cardiovascular disease is increased HDL-C, yet meats will also increase HDL-C.  Dairy and meats also have a similar nutrient profile: both are rich in vitamin K2, saturated fat as well as other nutrients less related to cardiovascular disease.  In section 3.1 the guidelines have said saturated fat increases risk factors for heart disease, but dairy has a greater proportion of saturated fat compared to meats.  It would be logical to think that with similar nutrient profiles these foods will exert similar effects, but that is not the outcome of epidemiological evidence.

“The evidence is difficult to interpret because of widely varying definitions of ‘meat’. Some studies include only unprocessed red meat. Others may include some or all of a variety of processed meats, including smoked, salted and chemically preserved foods, with meat within dishes such as pizza, lasagna or casseroles variously included or excluded.”

This poses a problem.  Meats are often grouped with processed meat and are often found is foods like pizza.  When we find poor health outcomes associated with pizza is it meat is the pizza, which only makes up a small part?  Or is it the white flour, the hydrogenated or deodorised vegetable oils and the chips and sugary drinks that pizza is paired with?

“As with other areas of diet and disease risk, an individual’s dietary pattern may be more relevant than a direct effect from a single component”

With guidelines such as these suggesting meats increase the risk of some disease and that plant-based foods are suitable alternatives, it’s not surprising that some people consider meats to be unhealthy.  Some health conscious people become vegetarian for that reason.  Some studies find data that suggests vegetarians live longer [40], but this is heavily confounded by health conscious people becoming vegetarians and eating more whole, unprocessed food [41].  Similarly, intervention trials based on Mediterranean diets may reduce red meat, but also emphasise whole, unprocessed food.  Multifactorial interventions such as Mediterranean diets cannot be used to draw conclusions about the health outcomes from any one food or nutrient.  The only conclusion that can be drawn is that the intervention improves health outcomes compared to the control group.  Once again, does removing meats from the diet improve health?  Or is this confounded by reducing processed and junk foods, thereby improving food quality in the diet?  This is supported by health conscious omnivores having similar rates of mortality than pragmatic vegetarians (health conscious people who decide to become vegetarian), and that pragmatic vegetarians are healthier than ideological vegetarians (removing meats for political or ethical reasons) [42].


Animal protein is superior to plant protein because of its high bioavailability, optimal amino acid ratios and the presence of amino acid derivatives.  Animal foods are rich in several nutrients plant foods are low in and vice versa.  Some of these nutrients are fat soluble, such as long chain omega 3’s, conjugated linoleic acids, coenzyme Q10 and vitamins A and K2.  Animal fats tend to be high in chemically stable SFA and MUFA, and low in the highly reactive PUFAs (see response to section 3.1).  So in order to consume a nourishing and low oxidative stress diet – animal protein and fat should make up most of the total protein and fat consumed.

Meats have been associated with poor health outcomes, but this may be due to other factors.  The guidelines express the limitations of the data by acknowledging the how various definitions of meat may combine red and processed meat, how unhealthy foods used to measure meat consumption and how health conscious people restricting meat is a strong confounding variable in epidemiology.

Sunday, February 5, 2012

DGA 2011 - Dairy and Alternatives

2.5 Milk, yoghurt, cheese and/or alternatives (mostly reduced fat)

Vitamin D - 1 µg = 40 IU
25-Hydroxyvitamin D - 2.5 nmol/l = 1 ng/ml

Full Fat or Low Fat?

“It is probable that the consumption of at least two servings per day of dairy foods (milk, cheese and yoghurt) is associated with reduced risk of ischemic heart disease and myocardial infarction (Grade B, Section 5.3 in Evidence Report [14]) [376].”

“It is probable that the consumption of two or more servings of dairy foods per day (milk, cheese and yoghurt) is associated with reduced risk of stroke (Grade B, Section 5.4 in Evidence Report [14]) [376, 377] particularly reduced fat varieties.”

The ‘particularly reduced fat varieties’ was not the conclusion of either paper.  The 376 paper did not want to make conclusive recommendations on the differences in health between whole milk and reduced fat milk for the following reason:

“Nevertheless, persons who choose to drink fat-reduced milks will almost certainly have adopted other “healthy’ behaviours, and these will undoubtedly be responsible for further confounding. These other factors cannot all be known, but they will be responsible for biases, which cannot possibly be estimated or allowed for. No reasonable conclusions can therefore be based on these data and we refrain from conducting any kind of meta-analysis or summary statistics.”

Generally observational studies find people who drink reduced fat milk have lower odds ratios potentially confounded by other healthy behaviours.  They did mention one study that found:

“In another case-control study the odds ratio for MI were significantly reduced (0.36; 0.13, 0.99) in subjects within the top quartile of adipose tissue C15.0 levels.”

The 15:0 fatty acid is found in ruminants and so can be used as a marker of dairy fat (and ruminant fat) consumption.  The more dairy fat consumed the more 15:0 will likely end up in adipose tissue.  Those in the top quartile would have likely consumed more full fat dairy and whole milk.  This study at least suggests that dairy fat is inversely associated with myocardial infarctions.  The authors conclude with:

“In the absence of evidence from large randomised trials the statement of German and Dillard is therefore most apposite: “Such hypotheses (about fat-reduced milks) are the basis of sound scientific debate; however they are not the basis of sound public health policy’.”

The authors of the 377 paper conclude by saying:

“We observed no statistically significant associations in this large cohort between intake of total fat, specific types of fat, or cholesterol and risk of ischaemic, haemorrhagic, or total stroke. In addition, consumption of red meats, high fat dairy products, nuts, or eggs did not seem to be related to risk of stroke.”

It appears the dietary guidelines have misrepresented the studies they cite.  One study concluded by saying that without RCTs, recommending reduced fat milk is not the basis of sound public health policy, the other found total fat, types of fat and high fat dairy to not be related to the risk of stroke.

“The proportion of total fat and saturated fat content in some milk, cheese and yoghurts has led to the recommendation that reduced fat varieties should be chosen on most occasions.”

The guidelines in section 3.1 to recommend reducing saturated and total fat consumption, but the cited studies don’t support either recommendation.  Instead they find low carbohydrate, high fat diets are equal to or better than low fat, high carbohydrate diets for weight loss [1] and that dietary fat improves blood lipids relative to carbohydrate [2].  Saturated fatty acids are also not associated with cardiovascular disease [3] and are inversely associated with atherosclerosis [4].

“Two proposed mechanisms link the consumption of milk, yoghurt and cheese products with a reduction in cardiovascular risk. Firstly, the consumption of milk, yoghurt and cheese products has been linked to an increase in the levels of high density lipoprotein (HDL) cholesterol.”

This proposed mechanism is out of touch with the recommendations.  HDL-C is increased by dietary fat, especially saturated fat, relative to carbohydrate [2].  Therefore the increase in HDL-C, the proposed mechanism by which dairy reduces the risk of cardiovascular disease, is proportional to the total dairy fat consumed.

Full fat dairy may be preferable to reduced fat dairy as dairy fat is rich in fat soluble vitamins such as A, D and K2, and also contains health promoting fats such as butyric acid (4:0) and other short to medium chain fats, long chain omega 3 PUFA and healthy trans-fats such as vaccenic acid and conjugated linoleic acids.

Butyric acid comprises 3-4% of dairy fat and short and medium chain fats comprise roughly 11-12% of dairy fat [5].  Butyric acid is likely to be the main reason why soluble fibre is healthy.  (Those values may be roughly 50% higher if grass-fed, which is based on assumptions that dairy in the USDA database are from grain-fed cows, the milk from the sheep and goats are grass-fed and the physiology between the species is not significantly different).  Soluble fibre is fermented by bacteria in the colon and butyric acid is a by-product, the butyric acid then nourishes the colon as its primary fuel source.  Butyric acid decreases inflammation [6] intestinal permeability [7] and is preventative against weight gain and insulin resistance [8].  Short and medium chain fats (4-12 carbon length) are more likely to be converted into ketones.  Ketogenic diets increase mitochondrial biogenesis [9], which may be therapeutic for a number of metabolic, age-related, neurodegenerative and psychiatric diseases [10].  Medium chain fats may also aid in weight loss by increasing fat metabolism and thermogenesis [11].  (This evidence also supports the consumption of coconut-based foods, which is the richest source of medium chain triglycerides).  Long chain omega 3 PUFA don’t need a mention and I’ve discussed the healthy trans-fats previous.

Vitamin K2 is an underappreciated and relatively unheard of essential fat-soluble vitamin.  Fatty meats, organ meats, eggs and cheese are the best sources and are in the form of menaquinone-4, vitamin K2 can also be found in some fermented foods as menaquinone-7 [12].  Vitamin K2 is responsible for activating vitamin A and vitamin D dependent proteins [13], moving calcium from soft tissue such as artery walls to use it in mineralising bones [14], which makes it effective at preventing both cardiovascular disease [15] and osteoporosis [16], as well as other functions in several organs such as the brain, pancreas and salivary glands [17].  The health benefits from vitamin K2 are not seen with vitamin K1 (phylloquinone) consumption [15].  Humans do not convert vitamin K1 into K2 and the K2 from intestinal bacteria is barely absorbed.

As a fat-soluble vitamin its intake depends on the quantity of fat consumed.  Americans, like Australians, have been told to reduce animal fat and cholesterol for decades.  With the saturated fat/cholesterol phobia we have cut the fat off our meat, stopped eating organ meats, choose reduced fat dairy and throw out the yolks, which leaves cheese.  Cheese is negatively associated with CHD and is a major source of K2 our diet [15].  This is probably because it’s now our major source of animal fat.  Even reduced fat cheddar cheese would have a fair amount of K2 are it is roughly 16-24% fat by weight.  Vitamin K2 is one factor that can explain why dairy is inversely associated with cardiovascular disease.  What’s also interesting is that vitamin K2 must have a very beneficial effect on health to overcome confounding variables related to its consumption and unhealthy lifestyles.

“In contrast to phylloquinone, intake of menaquinone (mainly MK-4 from eggs and meat, and MK-8 and MK-9 from cheese) is not related to a healthy lifestyle or diet, which makes it unlikely that the observed reduction in coronary risk is due to confounding.” [15]

The Value of Dairy, Calcium and Vitamin D

Despite being concerned with whole food rather than nutrients, the guidelines focus almost exclusively on calcium in regards to dairy consumption.  This is demonstrated by:

“The traditional nutritional rationale for the inclusion of dairy foods such as milk, cheese and yoghurt is their high calcium content and the positive relationship between calcium and bone mass.”

“Alternatives to milk, cheese and yoghurt include calcium-enriched legume/bean milk products such as calcium enriched soy drinks.”

However, these alternatives to dairy foods are very different nutritionally.  Calcium enriched legume milk products may compare well on calcium, but they likely do not contain vitamins A, B12, D and K2, long chain omega 3’s and conjugated linoleic acids [5].  As discussed above vitamin K2 is an essential nutrient and very useful in preventing arterial calcification and osteoporosis.  Without vitamin K2, calcium enriched foods may similarly increase the risk of myocardial infarction as calcium supplements do.

If calcium enriched foods aren’t an alternative to dairy, then what is?  There are people who have lactose intolerance or allergic reactions to dairy proteins.  Dairy is included in the diet mostly for calcium, but there are more factors involved in calcium balance and bone mineral density than just dietary calcium.

Vitamin D greatly increases calcium absorption.  Calcium absorption is 65% higher when vitamin D blood levels are at 86.5 nmol/l, relative to 50 nmol/l.  Both vitamin D blood levels are in the reference range, but the authors consider this to be a sub-optimal range for calcium absorption, instead they suggest the reference range for 25-hydroxyvitamin D should be 80-90 nmol/l [18].  If a young or middle aged adult increased their calcium absorption by 65% the RDI could effectively by lowered from 1,000 mg to 606 mg for them.  Even without an emphasis on vitamin D our calcium requirement for men and women may be 741 mg [19].  This now means increasing vitamin D could reduce the requirement down to 449 mg (~450 mg).  Now consider that 2.8% of New Zealanders aged 15 and over have a vitamin D deficiency (17.5 nmol/l) and 27.6% had a vitamin D insufficiency (37.5 nmol/l), or that the Geelong Osteoporosis study found the average vitamin D levels in summer were 70 nmol/l and in winter were 56 nmol/l [20].  Even 70 is below the current 75 nmol/l cut-off for vitamin D sufficiency.  Not many Australians have ideal vitamin D blood levels if ideal is 80-90 nmol/l (perhaps 80 in winter and 90 nmol/l in summer).  What is evident is that Australians and New Zealanders have a large room for improvement in increasing calcium absorption.

We may have low vitamin D blood levels because the adequate intake (AI) is too low.  The AI is 5 µg for people 50 years and younger, 10 µg for people aged 51-70 and 15 µg for people aged 70+.  The upper level is 25 µg for infants and 80 µg for everyone else [20].  The AI may be good enough to prevent rickets but isn’t high enough to achieve optimal calcium absorption.  One study found most people were able to reach 100 nmol/l with about 100 µg of vitamin D and 240 µg was needed for 97.5% of the population to have a vitamin D blood level of 100 nmol/l, while not toxicity occurred from vitamin D intakes of 250 µg [21].  The intake of vitamin D needed for 100 nmol/l of 25-hydroxyvitamin D varies greatly between individuals and depends on factors such as body weight [22].  Vitamin D toxicity occurs when 25-hydroxyvitamin is over 200 nmol/l, which requires an intake of 1,000 µg per day [23].  Maximum vitamin D intake from sunlight is only 500 µg per day [23], therefore it’s unlikely that vitamin D toxicity could occur naturally, unless one is either vitamin A or K2 deficient [24].

Current sun exposure guidelines may be too conservative and largely contribute to the widespread low levels of 25-hydroxyvitamin D in Australians.  Vitamin D has potent anti-cancer effects.  UVA is responsible for skin cancers and depletes vitamin D, while UVB is used to synthesise vitamin D.  UVA is around all day and can penetrate glass, clothing and sunscreen; UVB is only around while the UV index is greater than 3 and does not have the same penetrance [25].  Taking all of this into account it seems that the UVA:UVB ratio is predictive of skin cancer and is best kept low.  Sufficient midday sun exposure to as much of the body as possible to meet the vitamin D targets outlined above, while minimising sunscreen, and sunlight exposure through windows and clothing will keep the UVA:UVB ratio low, ensure good vitamin D levels and reduce the risk of osteoporosis and skin cancer and other forms of cancer.

The authors of the 376 study dismiss the idea that because our hunter-gatherer ancestors had strong bones without consuming dairy, therefore we don’t need to consume dairy either.  They dismiss it on the basis that our diet has changed.  Indeed it has changed.  Most Australians consume a highly refined diet largely devoid of key nutrients for bone health (such as vitamin K2), have poor vitamin D levels from limited sun exposure and we now also consume a grain-based diet, which reduces calcium absorption from phytic acid and compromises vitamin D metabolism [26].  A return to a diet and sun exposure based on our hunter-gatherer ancestors could produce strong bones without needing dairy.

An alternative to dairy would therefore include optimal vitamin D from mostly sunlight (to enhance calcium absorption) and other foods that have similar nutrients as dairy foods do, while assuming one consumes the recommended amount of fruit and vegetables.

“Milk, cheese and yoghurt are a good source of many nutrients, including calcium, protein, iodine, vitamin A, vitamin D, riboflavin, vitamin B12 and zinc.”

Foods that are food sources in these nutrients (except calcium) are other animal-based foods, such as meat, fish (including shellfish), eggs and organs.  All have highly bioavailable protein, with a complete amino acid profile and include nutrients derived from amino acids such as carnitine, carnosine, creatine, glutathione and taurine.  Fish, eggs and organs are good sources of iodine, eggs and liver are good sources of vitamin A.  All are good sources of riboflavin, vitamin B12 and zinc as well as vitamin B5, K2, selenium and long chain omega 3’s (which dairy is also a good source of).  Meat from ruminant animals is also a good source of the healthy trans-fats, vaccenic acid and conjugated linoleic acids.  Many of the nutrients listed above are found abundantly in other animal foods and animal foods are often the only good or unique source of them [5].

This suggests that an alternative to dairy is not calcium-enriched legume milks, but rather animal foods and ideal levels of vitamin D.  Although if dairy is replaced with animal foods and vitamin D in the context of a highly refined diet low calcium could become a problem, but in a highly refined diet health problems are ubiquitous anyway.


Raising HDL-C and vitamin K2 are some mechanisms that may explain how dairy foods decrease the risk of cardiovascular disease.  These mechanisms are proportional to the amount of dairy fat consumed.  Dairy fat contains nutrients such as vitamins A, D, K2 and healthy fats such as butyric acid, other short/medium chain fats, long chain omega 3’s, vaccenic acid and conjugated linoleic acids.  Therefore full fat dairy should be chosen to maximise health benefits when consuming dairy.

80-90 nmol/l, or even 100 nmol/l, are ideal blood levels of 25-hydroxyvitamin D for calcium absorption, as well as other health benefits.  240 µg is needed for 97.5% of the population to achieve a blood level of 100 nmol/l, although most people only need 100 µg and this depends on factors such as body weight.  The AI for vitamin D should be changed to 100 µg and an RDI should be made at 240 µg.  The recommendations regarding sun exposure should be changed to increase noon sun exposure to as much of the body as possible to promote health and not be responsible for skin cancers.

Alternatives to dairy should only include other animal foods along with an ideal blood level of 25-hydroxyvitamin D (80-90 or 100 nmol/l).  Animal-based foods are rich sources of the nutrients dairy foods are also rich in.  Many of the nutrients are only found adequately in animal-based foods and some nutrients are exclusive to animal foods.  This makes animal-based foods a better alternative to dairy than plant-based dairy substitutes.