The Cheat Codes for Optimal Nutrition, Satiety and Health

by

Ever wished you could cut through the noise and bias to understand what you should eat, tailored to your context and goals? 

I know I have!

I’ve been on a quest over the past 20 years to quantify nutrition to help my family and my community identify foods and meals that provide their bodies with the nutrients required from the foods they eat.  After nearly seven years of number-crunching and refining systems, I think I might be getting pretty close! 

This article showcases my findings based on the analysis of one hundred and fifty thousand days of food diary data from forty thousand people we collected from our Optimising Nutrition Community. 

The information in this article will empower you to cut through nutrition bias and dogma that dominates the nutrition space to give your body what it needs to be satisfied and thrive!

Contents

The Nutrient Leverage Hypothesis

Multivariate linear regression analysis of 111,906 days of food diaries from forty thousand people has enabled us to determine the most statistically significant factors that align with greater satiety and eating less. 

The most critical nutrients identified in this analysis were:

  1. Protein (%),
  2. Potassium,
  3. Fibre,
  4. Sodium,
  5. Vitamin B5, and
  6. Folate. 

While protein leverage is widely accepted, all nutrients appear to have a nutrient leverage effect. 

  • Our cravings increase, and we keep on eating until we get the nutrients (not the amino acids) we require from our food. 
  • Prioritising nutrient-dense foods can satisfy our cravings without consuming excess energy. 

The regression analysis of this extensive dataset across a wide range of dietary approaches provides correlation coefficients that allow us to calculate the effect of each nutrient and estimate the daily calorie intake of any food or meal. 

Imagine you were locked in a room for a week and had unlimited access to only one food or meal.  This analysis enables us to predict how much of a particular food or meal an average person would likely eat. 

So, if your goal is to satisfy your cravings, eat less and lose weight, you should prioritise foods and meals with a higher satiety index score.   

The Most Satiating Nutrient Dense Foods

Before we get into the detail, let’s look at the results you can use.  The chart below shows the estimated calorie intake vs diet quality score for a range of popular foods. 

  • In the bottom right, we have refined oils containing negligible amounts of protein and nutrients but plenty of energy.
  • In the top left, we have non-starchy vegetables and lean seafood that contain a lot of nutrients with minimal energy, which are much harder to overeat.

The vertical axis is our Diet Quality Score.  As you can see, there is some correlation between how much we eat and diet quality, but it’s not a direct relationship. 

To dive into the detail, you can check out the interactive Tableau version of this chart here.  For best results, use a computer—not your phone screen! 

You can also access our full suite of nutrient focussed food lists tailored to a range of goals and preferences here.  These are a great way to start your Nutritional Optimisation journey.    

Viewing all food on a scale of satiety vs nutrient density frees us from thinking in terms of:

  • plants vs animal-based foods,
  • low-carb vs low-fat, and
  • avoiding any ‘bad’ things in food. 

Rather than restricting calories or avoiding certain foods, this approach changes someone’s focus to simply seeking the nutrients we require from food.

Plant-Based Foods

The chart below shows the broad spectrum of plant-based foods.  Some of the foods that contain the most nutrients per calorie are plant-based.  But at the same time, some of the least nutritious and most processed foods are a conglomerate of refined plant-based ingredients. 

  • To maximise satiety, you should aim to consume more foods to the left of this chart. 
  • To maximise nutrient density, you can aim for foods towards the top. 

You can also check out the interactive Tableau version of this chart here.

Animal-Based Foods

It’s probably not going to be a surprise that when we look at animal-based foods, organ meats like liver and kidney rank well in terms of nutrient density and satiety. 

Check out the detail in the interactive Tableau version of this chart here.

Seafood

While not cheap, lean seafood like crab, crayfish, cod, and oysters are nutrient-dense and satiating.  Interestingly, caviar is nutritious but not as satiating because it contains less protein.   

You can see the interactive Tableau version of this chart here.

NutriBooster Recipes

But we usually don’t eat single foods.  We can also apply the updated satiety index formula to meals and recipes.  

The chart below shows forecast calorie intake vs diet quality score for our six hundred NutriBooster Recipes. 

Again, you can view the interactive Tableau version of this chart here.

The forecast daily calorie intake is based on the average population of forty thousand Optimisers.  Your actual intake will vary based on your size and activity levels. 

However, we can convert the estimated calories into a satiety index enables us to rank foods and meals from 0 to 100 based on how likely they are to keep you feeling full with fewer calories, as shown in the final chart below.

Click here to dive into the detail of all our NutriBooster recipes.  

So now you know which foods and meals will keep you full and give you the nutrients your body requires. 

If you’re ready to get on with optimising your diet, you can:

Read on to dig a little deeper to understand the analysis behind this. 

Nutrient Optimiser Data

Five years ago, we created Nutrient Optimiser to help people dial in their nutrients in our Macros Masterclasses and Micros Masterclasses.  Many others have also taken our free 7-Day Nutrient Clarity Challenge to identify the nutritional gaps in their diet. 

So, this data represents people with a wide range of goals.  Some are trying to lose weight; others are trying to improve their nutrients, but most people are simply assessing the macro and micro profile of their current diet. 

Excitingly, as our database grows, we get a clearer picture of how our macro and micronutrient intakes align with how much we eat.  We now have one hundred and fifty thousand days of food logging data from forty thousand Optimisers, representing how people eat worldwide. 

If you’d like ‘donate’ your Cronometer data to help us improve this analysis in the future and get an analysis of your current diet, you can click here to take our 7-Day Free Nutrient Clarity Challenge. 

Leverage Effect of Individual Nutrients

To compare nutrient density across all energy intakes, we normalised the nutrient intakes to nutrients per 2000 calories (or per cent of calories for macronutrients).  This allows us to plot satiety response curves like the one for potassium shown below. 

As with many nutrients, we can see that people who get more potassium per calorie tend to eat fewer calories.  But the satiety response and our cravings start to taper off once we get enough of that nutrient.

In terms of macronutrients, we found that:

  • fat and carbs provide energy, and eating a more significant percentage of them aligns with eating more overall, meanwhile
  • a higher protein % and a greater fibre intake align with eating less. 

We found that getting more of each essential nutrient per calorie for the micronutrients aligns with eating fewer calories.  In other words, packing more essential nutrients into our daily energy intake and increasing nutrient density aligns with eating less. 

While we have data on 11 amino acids, 12 vitamins, ten minerals, and two essential fatty acids, the obvious question is: which ones make the most significant difference when all the nutrients are considered together?  

This is where the multivariate linear regression analysis comes in.  This analysis enables us to ‘listen to our cravings’ to identify which nutrients we need more of and to what degree.   

Multivariate Linear Regression Analysis

Multivariate regression analysis (MVA) is a statistical analysis technique commonly used to understand which factors have the most significant impact on complex systems with multiple variables.  

The MVA can tell us which nutrients impact satiety most significantly.  If a particular nutrient was found to be statistically insignificant, it was eliminated from the analysis, so we focus only on the remaining nutrients. 

Limitations of the Multivariate Analysis

Unfortunately, nutritional data is not always accurate or complete. 

While some nutrition databases have data for all essential micronutrients, other food entries may only have data for the macronutrients.  Say a person eats twenty different foods in a day.  Some of them may have complete data for all the essential nutrients, some may just have good data for aminos, and others for vitamins.  So, the analysis may be biased toward the nutrients we have more comprehensive data for. 

Despite this, the MVA still shows a statistically significant nutrient leverage effect for many essential nutrients.  In the future, as we get more data from Optimisers over the coming years, we anticipate that even more may show up as statistically significant in our complex food matrix.  

Macronutrients

To dive into our analysis, let’s first look at the multivariate analysis of the macronutrients (i.e., protein, carbs, fat and fibre).   The chart below shows the percentage of protein, net carbs, fat and fibre vs calorie intake. 

  • We can see that higher protein (red line) and fibre (brown line) align with lower energy intakes.  Increasing fat tends to align with higher calorie intake. 
  • For non-fibre carbohydrates, we see a reduction in calories when we move from about 45% to 15%.  However, Towards the right, we see that a very high-carb, low-fat diet tends to be more satiating than a similar blend of fat and carbs. 
  • Overall, consuming most of your calories from carbs and fat with less protein and fibre results in the greatest energy intake.

The table below shows the multivariate linear regression analysis results for macros alone.  Because protein % is a dominant factor that implicitly considers energy from fat and non-fibre carbs, we are left with only protein and fibre as the statistically significant variables. 

P-value15th85thcalories%
protein (%)019%44%524-34.8%
fibre (g/2000 cal)2.8E-2891144124-8.3%

The very low p-values show that protein and fibre are strongly associated with calorie intake.  Hence, we can see this relationship is not due to chance.  

The MVA also provides regression coefficients that enable us to calculate the estimated calorie intakes that align each variable.  The table shows the 15th percentile and 85th percentile intake for protein and fibre. 

Using these regression coefficients, we can calculate that:

  • Moving from lower protein (19%) to higher protein (44%) aligns with a significant 524 calorie reduction (35%) in calories. 
  • Similarly, moving from low-fibre (11g/2000 calories) to higher-fibre (44 g/2000 calories) aligns with an 8% reduction in calories. 

If you read no further, the main takeaway from this article is that dialling back your energy consumption from fat and non-fibre carbohydrates while prioritising protein and fibre from whole foods will increase your satiety and empower you to eat less. 

But to continue digging deeper into the micronutrients, keep reading. 

Amino Acids

Rather than simplistically thinking about macronutrients, I’ve been eager to understand the influence of individual micronutrients on how much we eat. 

The chart below shows the satiety response to the essential amino acids.  Similar to the reaction from protein %, getting more of each amino acid per calorie aligns with a lower calorie intake. 

The table below shows the results of the multivariate analysis when we only consider the amino acids.  We see that cysteine, methionine, and phenylalanine are statistically significant.   

 P-value15th85thcalories%
cysteine7.41E-680.662.46-173-11%
methionine9.8E-161.24.8-139-9%
phenylalanine3.82E-052.38.0-79-5%
total    -26%

The fact that we don’t see a statistically significantly satiety response to all the amino acids could be because:

  • we don’t have as intense cravings for the other amino acids, or more likely,
  • the data for amino acids is not accurately measured in all foods. 

While it’s interesting to think about individual amino acids, we don’t have consistently robust data.  Hence, it’s better to consider all amino acids bundled into protein as a macronutrient. 

Minerals

Next, let’s look at the satiety response to each essential mineral.  

We see that the macrominerals—or minerals we require in more significant amounts—like potassium, sodium, phosphorus, and calcium tend to have a more substantial satiety response.   In comparison, smaller trace minerals like copper, manganese, and zinc elicit a smaller satiety response. 

This could be because:

  • data limitation and inaccurate measurements amongst databases as macrominerals are more accurately measured;
  • we have lesser cravings for these minerals; or
  • supplements and fortified foods often contain trace minerals, and these products do not increase satiety. 

However, despite these limitations, when we analyse the minerals only, we still see a statistically significant satiety response for potassium, calcium, selenium, and phosphorus. 

mineral P-value15th85th%calories
potassium3.27E-24418696061-152-10%
calcium2.99E-1304571915-103-7%
selenium6.31E-14072313-95-6%
sodium6.62E-1061461.95218.7-76-5%
phosphorus2.45E-27775.22327.5-60-4%
total   -32%

Vitamins

Let’s look at vitamins next. Towards the right, we can see that we get larger quantities of vitamins B3 (niacin), B5 (pantothenic acid), and E.  In contrast, we only require small amounts of B12 (cobalamin) and B9 (folate).

When we run the MVA on only vitamins, we see that vitamins B3, B2, K1, and B5 have an independent statistical relationship with calories. 

vitamin P-value15th85thcalories%
niacin (B3)1.22E-2131353-174-11.2%
riboflavin (B2)2.88E-491.34.7-93-6.0%
vitamin K15.68E-8733958-67-4.3%
pantothenic acid (B5)4.93E-154.013.3-49-3.1%
total    -24.6%

Parameters that Align with Eating More? 

While we prefer to think about getting more nutrients per calorie, many people want to know what they should avoid in food.  A lot of nutrition advice simplistically focuses on what we should avoid, like sugar, saturated fat, cholesterol, etc.  

To clarify, the table below shows the MVA for non-essential nutrients that provide energy. 

 P-value15th85thcalories%
saturated fat2.29E-16420.262.328615.9%
monosaturated fat9.71E-11914.752.123713.2%
starch4.74E-1810.848.120511.4%
sugar2.97E-449.667.71287.1%
total    47.6%

While many people prefer to avoid one of these things, we see that saturated fat, monounsaturated fat, starch, and sugar align with eating more. 

While these nutrients provide energy in whole foods, ultra-processed foods tend to contain all of them in their refined form together to create ultra-profitable, hyper-palatable nutrient-poor foods.  

Bringing It All Together

Rather than looking at amino acids, vitamins, or minerals separately, this analysis becomes extremely useful when we look at the effects of all the nutrients together. 

Because there are several options and limitations depending on the data available, I’ve included a few scenarios. 

Top-Down Analysis: Maximum Data

This first table shows the results of the MVA when we consider all of the essential nutrients.  This analysis includes 111,897 days of data for which we have complete information on the essential nutrients. 

NutrientP-value15th85thCalories%
protein019%44%-486-31.6%
potassium2.18E-3819315915-72-4.7%
fibre7.96E-481144-70-4.5%
sodium1.14E-2814805076-41-2.7%
calcium1.69E-174691869-40-2.6%
pantothenic acid (B5)0.006415-18-1.1%
folate0.22167956-7-0.4%
total-47.6%

At the top of the table, protein % has the highest statistical significance.  Moving from 19% protein to 44% protein aligns with a substantial 32% reduction in calories.  However, while the effect of protein dominates in terms of satiety, we see an additional 16% reduction in calories from other statistically significant nutrients.   

The chart below shows the results of this analysis in terms of forecast calorie intake vs Diet Quality Score.  To dive into the detail, you can check out the interactive Tableau version of this chart here.

I’ve included a few other scenarios.  However, this first option uses the most data and only considers the nutrients that positively influenced satiety provided the best results. 

Top-Down and Bottom-Up Analysis

Unfortunately, we have fewer days of data to analyse if we want to consider factors that negatively influence satiety. 

The table below shows our analysis of 27,475 days of data.  This dataset contains information on the nutrients that contribute to energy consumption and the ones positively impacting satiety.  

NutrientP-value15th85thCalories%
protein  3.67E-20019%44%-427-23.5%
fibre  1.09E-241144-118-6.5%
sodium1.66E-2314805076-95-5.2%
potassium2.43E-0719315915-69-3.8%
riboflavin (B2)1.07E-051.335.03-61-3.4%
folate3.52E-05167956-52-2.8%
selenium1.30E-0672295-41-2.2%
monosaturated6.22E-071652543.0%
starch1.08E-340.849.6925.1%

In this MVA run, protein is still dominant, and consuming more protein was statistically significant for decreasing the number of calories someone ate.  However, in contrast to our first analysis, vitamin B2, folate, and selenium were also significant.  This suggests that a wide variety of vitamins and minerals may have a significant satiety impact depending on the context. 

At the bottom of the table, we see that reducing monounsaturated fat and starch will help reduce calorie intake.   

The chart below shows the ranking of foods using this scenario.  Overall, the results are similar.  The only real difference is that refined oils that contain more monounsaturated fat received negative rankings. 

Processed foods that contain a large proportion of energy from starch or monounsaturated fat should be avoided, particularly when combined.  However, simply paying more attention and focusing on the factors that positively influence satiety—like protein, vitamins, and minerals—tends to look after this and decrease your intake of foods containing this lethal combo innately. 

Top-Down Analysis: With Cholesterol

Cholesterol is not an essential nutrient, so I haven’t included it in the MVA runs above. 

But you may find it interesting that we always see an independent, statistically significant positive satiety response to foods that contain more cholesterol!  

As you can see in the chart below, we have a strong satiety response to foods that contain more cholesterol. The amount of cholesterol we consume is tiny (about two to ten calories per day), so it doesn’t contribute to an energy excess like monounsaturated fat and saturated fat.

Since we began listening to mainstream ‘heart healthy’, ‘fat-free’, and ‘low-cholesterol’ guidelines and relying on the ‘plant-based’ large-scale, industrialised agricultural food system in the 1950s, our cholesterol intake has declined obesity has climbed.

It was previously suggested we minimise cholesterol, but it was removed from the 2015 Dietary Guidelines as a nutrient of concern as it was found that there was a lack of evidence linking our intake of cholesterol with heart disease.

We also see a similar trend with sodium, which has a strong, consistent, and positive impact on satiety.  Interestingly, the decline in sodium consumption is most strongly correlated with our increased energy intake and the resulting obesity epidemic that has overtaken our population this past half-century. 

Avoiding cholesterol and sodium may cause us to eat more because our bodies crave these nutrients.  In their absence, our bodies send us in search of foods that contain more of them, and we eat more (calories) until we consume what we require.

The table below shows the MVA using the 34,042 days we have cholesterol data for.  Moving from low to higher cholesterol intakes aligns with a statistically significant reduction in calorie intake, amounting to a whopping 12%!

Nutrient P-value15th85thCalories%
protein (%)4.36E-22519%44%-364-20.5%
cholesterol6.89E-1442411137-214-12.1%
fibre1.62E-491144-153-8.6%
potassium2.13E-1919315915-97-5.5%
sodium2.93E-3014805076-95-5.4%
calcium1.59E-114691869-67-3.8%
total    -55.8%

The chart below shows the same forecast daily energy intake vs Diet Quality Score chart when cholesterol is included.  Interestingly, organ meats don’t shoot to the top of the ranking.  However, we do see some change in the rankings of the non-starchy vegetables that contain calcium and a little bit of shuffling of the fats in the bottom right corner. 

Top Down and Bottom Up: with Cholesterol

For completeness, the MVA run below shows the top-down and bottom-up analysis that considers cholesterol along with the essential nutrients and the nutrients that contribute energy to our diet. 

Nutrient P-value15th85thCalories%
protein  7.71E-12619%44%-355-19.5%
cholesterol  2.99E-722411137-192-10.6%
fibre  2.15E-501144-185-10.2%
potassium2.47E-1719315915-116-6.4%
sodium1.49E-2014805076-91-5.0%
folate1.29E-05167956-54-3.0%
selenium0.01472295-21-1.1%
riboflavin (B2)0.29115-13-0.7%
calcium0.404691869-10-0.5%
starch  2.01E-20150724.0%
monosaturated fat7.64E-161652904.9%

Selenium, calcium, and vitamin B2 still remain significant in this analysis, but their weight in impact decreases when other factors in the complex food matrix are considered.   

It’s interesting to note that saturated fat doesn’t correlate with increased calorie intake.  Instead, we see that consuming more cholesterol and sodium per calorie aligns with eating less and eating more starch, and ‘heart-healthy’ monounsaturated fats align with eating more.  

The results of this scenario are shown in the chart below.   However, other than seeing lard, tallow, and coconut oil ranking slightly above olive oil and avocado oil, we don’t see many major changes. 

I’m not sure the world is ready to accept that cholesterol is a positive nutrient we must prioritise to eat less!  Hence, it’s probably better to stick with the first model that excludes cholesterol so we can use the most available data. 

I don’t think we need to glorify cholesterol as another magical nutrient that we all should be drinking in unlimited quantities.  However, it’s high time we dropped it from the discussion about cholesterol as a ‘bad’ nutrient we need to be afraid of.  Doing so only leads us into the hands of Big Food, which wants to sell us ultra-processed refined ‘plant-based’ products (i.e., starch, sugar, and industrial seed oils blended with flavours and colourings). 

For more on cholesterol, see:

Dietary Cholesterol and Blood Cholesterol: Are They Related?

The Satiety Index

Our analysis enables us to identify foods and meals that provide more nutrients our bodies need to satisfy them without consuming excess energy.  Because the regression coefficients allow us to estimate the daily calorie intake for any food or meal, we can use them to rank all foods from 0 to 100 based on which ones are likely to provide the most vs least satiety. 

We have used this understanding to develop our updated satiety index to rank foods and meals.  For more details, see:

The good news is that you don’t have to adopt a new belief-based extreme dietary approach.  It doesn’t matter if you prefer to eat more plants, animal-based foods, or seafood; so long as you get the nutrients your body requires, you can eat whatever you want or like!

Some examples of high-satiety plant foods, animal foods, and seafood are shown below. 

For a complete list of the highest satiety foods in printable form, you can grab our bundle of nutrient-dense foods tailored for a range of goals and preferences here.  

Why Is Optimal Nutrition Important?

If you’ve made it this far, you may still be wondering how nutrients influence satiety or if Nutritional Optimisation is really important. 

I think you can look at this in three ways. 

  1. Nutrient leverage,
  2. Nutritional safety factors, and
  3. Copying the eating habits of successful people.

Before we wrap up, let’s quickly look at each.  

Nutrient Leverage

Professors Raubenheimer and Simpons’ Protein Leverage Hypothesis states that all organisms, including humans, crave protein and will continue eating until they get enough of it.  Hence, we will overconsume energy in our pursuit of protein if the only food we have available has a lower percentage of total calories from protein (i.e., protein %).

Our analysis indicates that there is not just a Protein Leverage Effect but a nutrient leverage effect. 

In other words, we will eat more energy until we obtain adequate amounts of all the nutrients we need.  Hence, we will be satisfied with fewer calories if we can pack more nutrients into our daily energy budget.

When we eat food that contains more nutrients, we appear to consume fewer calories.  Thus, it makes sense to pack more of the required nutrients into our daily energy budget.

Nutritional Safety Factor

In engineering, we apply a risk-based safety factor to manage unknown variables. 

You probably wouldn’t feel comfortable driving an overloaded truck over a bridge designed for the absolute minimum cost and weight without any safety factors, would you?  That’s why I would apply larger safety factors as a bridge designer to manage the unknowns so we could ensure that the bridge would not collapse during any foreseeable event.

Similarly, you can see striving for the Optimal Nutrient Intake as increasing the safety factor of your nutrition.  In line with Bruce Ames’ Triage theory, you are working to give your body plenty of all the nutrients you require rather than getting just enough of the essential nutrients to survive.  This ensures your body is well-prepared for anything you may face now or in the future.

Imitating Successful People

If you don’t buy into nutrient leverage or the nutritional safety factor argument, you can simply view Nutritional Optimisation as imitating the habits of successful, healthy people.

If you see someone who is fit and full of life, you often want to know what they’re eating and their workout routine so you can imitate them. 

Right?  

Through our analysis of the eating habits of forty thousand people worldwide, we have quantified the nutritional parameters that align with higher satiety.  You can think of these as data-driven cheat codes for nutrition.

Our data-centric analysis enables us to cut through the noise and from endless arguments centred around plant-based vs carnivore, low-carb vs low-fat, the latest ‘anti’-nutrients to avoid or the latest epidemiological study.  

Your body doesn’t care where it gets its nutrients from; it just needs enough of each of them!  Once you get everything you need within your daily energy budget, anything ‘bad’ things become irrelevant.

By prioritising nutrient density, you can optimise your health, satiety, and vitality with the foods and meals you enjoy available in your corner of the world!

Which Nutrients Do YOU Need More Of? 

While this large dataset gives us highly accurate information about how most people eat most of the time, one caveat is that the perfect nutritional prescription for you depends on how you currently eat.  In other words, you need to focus on what you currently aren’t getting enough of—not what you’re already getting plenty of, or not what someone else needs! 

While most people benefit from more protein and potassium, you may already be getting plenty.  Hence, your cravings for that nutrient will be lower, and you might do better if you direct your appetite towards other nutrients you may not get enough of. 

While the rankings shown in the charts above will satisfy most peoples’ cravings, you are still unique.  Ideally, you want to identify the cluster of nutrients you are currently getting less of, and hence the ones you need to prioritise. 

We’ve designed our suite of nutrient-dense food lists and NutriBooster recipes to make it as easy as possible for people to get started on their journey of Nutritional Optimisation.  However, if you want to make a little more effort to find the nutrients you need to prioritise the foods and meals that contain them, you can use our Free 7-Day Nutrient Clarity Challenge. 

Later, you can take our Micros Masterclass to take your nutrient density to the next level.  In just four weeks, you will learn how to optimise YOUR diet at the micronutrient level to give your body what it needs. 

More