Q: To achieve optimal health, diabetes control and longevity you should:
- Eat more fat to burn your body fat.
- Follow a Protein Sparing Modified Fast (PSMF).
- Eat more protein and lift heavy.
- Drink Bulletproof Coffee.
- Fast more.
- Decrease protein and eat more fat bombs.
- Take exogenous ketones, butter and MCT oil to boost your blood ketones.
- Eat only plants.
- Eat no plants.
- All of the above.
- None of the above.
- I give up! All you ‘diet gurus’ can’t agree. I’m going back to Maccas where things are simple!
Although many of these answers are contradictory, all are ‘correct’ depending on which low carb / keto group(s) you belong to. It can be confusing out there on the interwebs!
For the last two years I’ve been working to refine our ability to quantitatively define and optimise our food quality (a.k.a. nutrient density).
At the start of 2017, I developed the Nutrient Optimiser and have since run detailed macronutrient and micronutrients analyses for more than forty people, all with different starting points and with different goals.
With all the conflicting advice out there and my personal quest to manage diabetes while maximising nutrient density, I wondered what my nutrient analysis tools might be able to tell us about the relationship between macronutrients and micronutrients to provide some clarity to the circular debates that I see so often online.
I’m never sure where these articles will end up when I start the analysis. And this one is certainly interesting!
The analysis suggests that a nutrient dense diet is typically not low in protein. However just focusing on increasing protein won’t necessarily lead to a nutrient dense outcome.
We get a much better outcome when we focus on the harder-to-find micronutrients (i.e. vitamins, minerals and essential fatty acids). From there we can tweak the nutrient dense template to suit our goals (e.g. weight loss, diabetes control, muscle gain, athletic performance or therapeutic ketosis).
Let’s quickly look at what we mean by ‘nutrient density’ and how we can quantify it.
The chart below shows the nutrients provided by the 8,000 foods in the USDA database in terms of the percentage of the Daily Recommended Intake (DRI) if you ate just a little bit of all of them.
It’s easy to meet the recommended minimum intake of the micronutrients shown at the bottom of the chart (e.g. vitamin B12 and most of the amino acids) (at least if you are eating animal products).
However, you really have to go out of your way to get adequate amounts of the nutrients at the top of the chart (e.g. omega 3, vitamin D, choline, vitamin E, calcium, manganese and magnesium).
The most nutrient dense foods
The chart below shows the micronutrients provided by the most nutrient dense foods. When we focus on foods that contain more of the harder-to-find nutrients we can get a massive boost in all the micronutrients.
Why should we pursue a nutrient dense diet?
With adequate amounts of nutrients being provided by the food we eat there is a good chance we will be able to satisfy our cravings with less energy.
Obtaining adequate levels of all the micronutrients will ensure that we have what we need to drive our mitochondria at full power rather than limping along. We will feel energised and may find that our appetite turns off sooner and we will be less likely to overeat and get fat.          
The chart below shows a comparison of the most nutrient dense 10% of the foods available compared to all the foods in the USDA database. We get a significant improvement in our food quality by prioritising more nutrient dense foods.
Which nutrients do we need to worry about?
After a ton of trial and error and systems refinement (and some robust debates with Ray Cronise) I finally figured out that maximising nutrient density works best when we only focus on boosting the nutrients that are harder to obtain.
The nutrients listed below tend to be generally harder to get in adequate quantities:
- Alpha-linolenic acid
- EPA + DHA
- Vitamin D
- Vitamin E
- Pantothenic acid
Which nutrients are easier to find?
Listed below are the micronutrients that we don’t need to prioritise because they are fairly easy to get enough of:
- Vitamin B-6
- Vitamin C
- Vitamin A
- Vitamin B-12
- Vitamin K
I have intentionally left out all the amino acids (i.e. protein) from the prioritisation because, as you will see below, it’s easy to get enough protein when we focus on the vitamins, minerals and essential fatty acids.
Can you get too much of a good thing?
As a general rule it’s hard to get excess micronutrients from real food, but it is possible.
- While we can get more than thirty times the DRI for vitamin K from a nutrient dense diet there is no upper toxicity level of toxicity for vitamin K  from natural sources. However you can get too much menadione which is used as a vitamin K supplement. 
- We can get eighteen times the DRI for vitamin B12 from a nutrient dense diet, however again, there is no upper limit established for B12. 
- We can get seventeen times the DRI for Vitamin A from a nutrient dense diet. It is possible to get vitamin A toxicity, though again this typically occurs from supplementation.  There are some reports of Hypervitaminosis A from explorers gorging on polar bear liver, but this is not likely to be a common occurance. 
- We can get around twelve times the DRI for copper from a nutrient dense diet which is around the upper limit. Though these high levels are unlikely to occur without high liver consumption which is not common.
- A nutrient dense diet can provide around fifteen times the DRI for vitamin C however the upper limit is more than 20 times the DRI.  Excessive vitamin C supplementation usually causes diarrhea, so it’s largely a self limiting situation.
- A nutrient dense diet will provide around ten times the DRI for iron while the upper limit is set at around six times the DRI. Many women are iron deficient while many men have hemochromatosis which is excess iron storage. Liver, mushroom, seaweed and spices are the highest sources of iron. It’s useful to understand your current iron status to know whether you need more or less iron or should even be considering donating blood.
- It is quite easy to get more than the DRI for amino acids. While high protein diets do not cause kidney disease in healthy people there is no need to chase excess super high levels of protein. And just like liver, most people will struggle to eat excessive amounts.
So yes, it is possible to get excessive levels of some micronutrients, though generally not a concern unless you are eating a LOT of liver or supplementing with synthetic nutrients.
The chart below shows the nutrient profile of Amy who is following a zero carb diet with a lot of organ meats. While she is generally getting high levels of most nutrients, she is still not meeting the DRI for a number of vitamins and minerals that are typically found in plant foods (e.g. vitamin K1, calcium, manganese, vitamin E, magnesium and potassium).
At the other extreme we have David who is eating a plant based diet that has plenty of vitamins and minerals but less amino acids. He knows he needs to supplement with vitamin B12 and vitamin D which are hard to get from a purely plant based diet.
When it comes to nutrient density I often see arguments around whether or not the daily recommended intake levels are correct and whether they might vary for different people with different dietary approaches and whether or not nutrients from plant or animal based food are more bioavailable.
While I think these are definitely under researched areas I think these discussions are not so relevant when we’re orders of magnitude above or below the DRI values. We need to identify the full range of foods, from whatever source, that will provide the nutrients that we’re not getting enough. We can then choose from within those to suit our tastes and preferences. Our appetite can be a pretty good guide once we eliminate the processed hyper palatable nutrient poor foods that our willpower is no match for.
There is plenty of discussion about excess protein or excess calories. While it’s true that excess is typically not good, I think it’s more valuable to focus on eating foods that contain more of the nutrients that we are currently not getting enough of. When we’re eating nutrient dense whole foods we’re less likely to need to consciously worry about calories, protein, fat, carbs, sugar, fibre or whatever.
Is there any relationship between macronutrients and nutrient density?
While I don’t see a lot of discussion about nutrient density or food quality, there is seemingly endless debate in social media in low carb and keto circles around macronutrients. People are often very passionate about eating more or less protein, carbs, fat and fibre.
Perhaps this is because macronutrients are reasonably easy to track and understand. Or maybe it is because the previous approach hasn’t worked, so they swing to the other extreme.
We’ve been told for so long that fat is bad and now people are realising that it’s not as bad as they were told, so they swing to the other extreme. Now fat can do no wrong.
Meanwhile, there are plenty of people who stick to fat being bad and wanting to avoid it.
Different people have different perspectives on the multifaceted topic of nutrition.
But is there really any value gained by focusing on primarily on macronutrients? Will it improve our food quality or the adequacy of the various essential micronutrients?
To understand whether there is any useful relationship between the various macros and micronutrient adequacy I have plotted the various macronutrients versus the nutrient density score for the 8,000 foods in the USDA foods database.
Note: In this analysis a high nutrient density score means that a particular food has a relatively large amount of the harder-to-find nutrients listed above.
Protein versus nutrients density
There is a lot of debate about protein and whether we should be getting more or less of it.
The chart below shows the nutrient density score for the harder-to-find vitamins, minerals and essential fatty acids vs protein (%).
Although amino acids have not included in the nutrient density score it appears that the more nutrient dense foods have more protein. Conversely, foods with less protein have less of the nutrients that are harder to find.
It seems that if we avoid protein we will end up with less nutrients overall. While if we focus on getting the nutrients that are harder to find we will get enough protein.
However, as they say, correlation does not equal causation. There is a lot of scatter in this chart. In this case the correlation (R2) of this relationship is 0.31.
This analysis makes me wonder if the studies that the benefits from increased protein are not at least in part from, not just getting adequate amino acids, but the increased levels of the other micronutrients that often come along with protein.
It’s hard to separate good nutrition and protein.
Fat versus nutrient density
The chart below shows the nutrient density score versus the percentage of calories from fat.
The first thing to point out here is that there is a massive amount of scatter and a low degree of correlation between fat and micronutrients (R2 = 0.06).
However, it does seem that very high fat foods contain less of the harder-to-find nutrients.
Meanwhile at the other extreme very low fat foods can either be nutrient poor (e.g. sugar and processed grains which would be at the bottom left of this chart) or very nutrient dense (e.g. non-starchy vegetables which would be at the top right of this chart).
If we run a trend line through all these foods we see that the highest nutrient density occurs at around 30% calories from fat.
The reality is that not many people live primarily on high nutrient density low fat foods at the top left corner of this chart. People avoiding fat will often slip into the bottom left of this chart and resort to the low fat processed grains and sugars to get enough energy to get through the day.
Sugar versus nutrient density
There is currently a lot of focus on sugar as the primary culprit for our poor health. Gary Taubes and Damon Gameau are down on sugar while Robert Lustig is leading the charge against fructose or fruit sugar.
This analysis suggests that foods with more sugar have a poorer nutrient density, though it’s hard to make sense of this unless we differentiate between added refined sugar and naturally occurring sugar in plant based foods that come with a ton of other nutrients. However, low sugar content does not necessarily guarantee excellent nutrient density.
Energy density versus nutrient density
Energy density is the amount of energy we get per gram of food.
Minimally processed foods contain more water and fibre and thus have a lower energy density but also tend to have a higher nutrient density.
Meanwhile, processed foods that are shelf stable and easy to transport typically have less water and fibre and more preservatives.
While lower energy density foods have a higher nutrient density, most people won’t survive long on a diet of only lettuce, broccoli and celery. They will need some more energy dense foods to survive.
However, if you are looking to lose weight in a hurry while still getting the nutrients you need, focusing on lower energy density foods might not be a bad place to start.
Most people agree that eating more veggies will be better for their health, but the unfortunate reality is that it takes some time and money to prepare the food yourself rather than reaching for a quick and cheap energy hit with minimal effort.
Net carbs versus nutrient density
Foods with more digestible carbohydrates typically have a lower nutrient density.
However, simply going low carb doesn’t guarantee that we maximise nutrient density There is a range of high and low nutrient density foods at the low carb end.
Whether or not you carbs are nutrient dense will likely depend more on whether they are highly processed or in their natural form, and will likely make a bigger contribution to their nutrient density than the quantity of carbs.
Higher fibre foods contain more nutrients. However, we can’t just add fibre supplements to maximise nutrient density. Plant based whole foods that also happen to have heaps of fibre that provide us with more higher levels of nutrition.
The proportion of insulinogenic calories is the proportion of the food we eat that requires insulin to metabolise.
On the right hand side of the chart, highly processed foods with minimal protein and fat typically don’t provide a lot of the harder-to-find nutrients.
Meanwhile on the left hand side of the chart, foods with minimal fibre, carbs and protein are also less nutritious.
If we plot a trendline it appears that the maximum nutrient density occurs at around 50% insulinogenic calories.
If you are already insulin resistant you may want to steer your dietary ship to the left with a lower insulin load diet to the point that your pancreas can keep up and maintain normal blood sugars. Meanwhile if you’re fit and insulin sensitive you will be able to have more leeway when it comes to macros and insulin load.
So what to make of all this? Which of these parameters has the best correlation with food quality or nutrient density? The table below shows the various parameters sorted by their correlation (R2) with their nutrient density score.
|protein||0.32||Nutrient dense foods tend to have more protein.|
|energy density||0.15||Lower energy density foods are typically more nutrient dense.|
|net carbs||0.12||Foods with more net carbs are typically less nutritious.|
|insulinogenic||0.11||Nutrient density peaks at around 50% insulinogenic calories. Extremes are not optimal.|
|fibre||0.08||High fibre foods are often more nutritious.|
|fat||0.07||Nutrient density peaks at around 30% fat.|
|sugar||0.04||High sugar content correlates with low nutrient density|
It seems that if we want to optimise the quality of our diet we should:
- Focus on the foods that contain the harder-to-find nutrients.
- Not actively avoid protein.
- Chose lower energy density foods when we can.
- Avoid foods that are largely digestible carbs with minimal fibre (e.g. processed grains and sugars).
- Chose moderately insulinogenic foods without swinging to either extreme (though we should err on the less insulinogenic side if we already have diabetes).
Meanwhile, sugar, fat and fibre, aren’t spectacular predictors of nutrition.
chasing nutrients vs chasing macros
So, if protein is good, more is better, right? Bring me the bulk tub of protein powder!
Not so fast. It is important to understand the difference between emphasising:
- all nutrients,
- less insulinogenic foods, and
- harder-to-find nutrients.
Maximise all nutrients
The chart below shows what happens to the micronutrient profile when we simply maximise all nutrients.
The amino acids are through the roof (69% protein) because aminos are easy to find in our food system, but we’re still lacking in many of the harder to get nutrients.
If nutrient density correlates with protein then it makes some sense to prioritise protein. Doesn’t it?
The chart below shows what happens to the nutrient profile if we sort the USDA foods database by % protein. It seems that if we simply focus on protein we get a poor vitamin and mineral profile.
Minimising protein and maximising fat
Minimising protein and carbs while maximising fat is all the rage in the keto scene. Unfortunately, a very low insulin load diet is not a high nutrient density approach as we can see from the chart below. While we get adequate protein (15%), the vitamin and mineral profile is poor. With 80% of our energy coming from fat we are deficient in about half the micronutrients.
Perhaps a very high fat therapeutic ketogenic approach should be reserved for special circumstances and extra attention given to the nutrients you won’t be able to get from your food?
Prioritising the harder to find nutrients
The chart below shows the outcome when we focus the harder to find nutrients (excluding amino acids). We get adequate quantities of all the micronutrients and still plenty of protein.
Learnings from the Nutrient Optimiser analysis
It’s one thing to look theoretically in a database of individual foods. But it’s another to look at what people are eating in real life. Next, I’m going to share what I’ve learned from analysing a lot of different people’s food logs in the Nutrient Optimiser.
The nutrient density score
But first, I need to introduce you to the Nutrient Density Score.
The chart below shows Rhonda Patrick’s nutrient analysis. Rhonda’s diet is not particularly extreme in anything other than nutrients.
Rhonda would score 100% if she could achieve 200% of the DRI for the hardest to hardest to find lower half of the nutrients. However, because she doesn’t achieve 200% with all of the lesser scoring half of the nutrients she only gets a Nutrient Density Score of 81.3%.
For reference, if we add a little bit of all the foods in in the USDA database we would get a nutrient density score of 63% . The most nutrient dense 10% of the foods in the USDA database will give us a nutrient density score of 93%. Even Rhonda has some room for improvement.
By contrast, the chart below shows Patrick’s nutrient density score which comes in at only 21%. Patrick is following a very high fat keto approach even though his blood sugars are great and he doesn’t appear to be insulin resistant, just obese.
With so many of his micronutrients being nowhere near the DRI vales Patrick will need to eat a lot more of his current diet to meet the daily recommended intake for most of the nutrients.
There is a good chance that that Patrick will be craving more food to obtain the nutrients that he needs to get through the day. Even though he is trying to lose weight, he might end up overeating more calories using his current diet than if he spent a week eating with Rhonda.
The table below shows the nutrient density score for more than forty Nutrient Optimiser analyses that I’ve run to date along with:
- protein (g/kg LBM),
- protein (%)
- fat (%)
- fibre (%)
- net carbs (%).
I encourage you to click on each of the names below to review their nutrient analysis to see what they are and aren’t eating to get these scores.
|Name||score||protein (g/kg LBM)||protein||fat||fibre (%)||net carbs (%)|
|Andy Mant v3||77%||4.4||27%||53%||5%||15%|
|Alber Van Zyl||75%||1.0||15%||77%||2%||6%|
|Deb Pinsky Lambert v2||72%||1.2||31%||61%||3%||6%|
|Andy Mant v2||70%||3.0||26%||54%||6%||15%|
|Robin Reyes v3||69%||1.6||18%||67%||6%||8%|
|Ruth Jamieson v2||66%||1.6||18%||67%||6%||8%|
|Ruth Jamieson v1||57%||1.4||19%||65%||7%||9%|
|Andy Mant v1||34%||3.7||35%||54%||2%||9%|
|Robin Reyes v2||32%||1.6||21%||59%||4%||15%|
|Robin Reyes v1||23%||1.1||13%||50%||2%||35%|
|Patrick Butts v1||21%||0.8||18%||73%||4%||5%|
|Patrick Butts v2||20%||1.4||26%||66%||3%||6%|
In the charts below we’ll quickly look at the relationship between the macros and their nutrient score.
This chart shows the relationship between protein intake and each person’s nutrient density score. The average protein intake for this range of people following a low carb or keto diet is 2.1g/kg LBM or 23% of energy.
On the top left corner of the chart we have David who is following a plant based diet and intentionally getting lower levels of protein but also maximising vitamins and minerals from plant based foods.
On the bottom left we have a number of people following a therapeutic ketogenic diet targeting low protein and high fat.
As long as you are not trying to target low protein and high fat to generate higher blood ketones then it doesn’t seem to matter what your protein intake is. Most people get enough protein to support their activity levels.
The chart below shows the nutrient density score versus protein (%). Again, it seems that it’s hard to get high levels of nutrients if you are targeting minimal protein levels.
The story is similar with insulin load. Reducing the insulin load of your diet to the point that your blood sugars normalise is a great idea, but less is not necessarily better. We want to avoid really high insulin levels but not drive it so low that we don’t have enough nutrients to repair our muscles and organs.
High levels of fat do not guarantee high levels of nutrition.
It’s good to reduce the carbohydrate load of your diet to normalise your blood glucose levels, but again minimising is not necessarily the best idea and may be unnecessary if you are not managing diabetes.
Higher levels of isn’t necessarily bad either when it comes to nutrient density. On the top right of the chart we have David who is striving for a nutrient dense plant based diet with about 35% net carbs while for contrast we have Robin’s baseline junk food diet which also has about 35% net carbs which has about the same nutrient density score as the very high fat therapeutic keto dietary approaches on the bottom left of the chart.
Higher levels of fibre typically correlate with more nutrition (although you can get heaps of nutrients from shellfish and organ meats with minimal fibre intake).
- A nutrient dense diet is not low in protein; however focusing on protein won’t necessarily guarantee great nutrition.
- Foods with a lower energy density are often more nutrient dense. To maintain our body weight and growth we will need to add more energy dense foods (i.e. more non-fibre carb and / or fat). Meanwhile, dialling back the energy density and forcing your body to use your stored body fat can be a good strategy for weight loss.
- Reducing your carb intake or the insulin load of your diet can be useful if you are managing diabetes. However less is not necessarily better.
- For the most part ensuring you are getting the harder-to-find micronutrients will maximising your diet quality without going to macronutrient extremes.