Unlocking Satiety: The Breakthrough Algorithm That Decodes Your Cravings

Why do some foods satisfy you, while others leave you craving more—even when you’re full?

Using 872,925 days of real-world data, we’ve cracked the code on how different nutrients trigger hunger, fullness, and cravings. The result? A revolutionary satiety algorithm that reveals precisely how the foods you eat influence how much you eat.

Over the past few decades, most diet advice has focused on single factors, like cutting sugar, boosting fibre, or eating more protein.

But what if the secret isn’t about more or less of any single nutrient?

Once we understand that the problem is modern foods designed to hit our multiple bliss points, the solution becomes obvious: optimise your food choices to get more than the minimum amount we crave to trigger satiety on multiple levels.

In this article, you’ll discover:

The science behind why certain foods lead to overeating.
How our algorithm predicts your satiety with stunning accuracy.
How to move beyond the bliss points to satisfy your cravings more efficiently.

Contents

The Ultimate Goal
What Drives Your Hunger? The Key Factors of Satiety
The Science of Cravings: Understand Your Nutrient Bliss Point
What About the Rest? The Bigger Picture of Satiety
Data Quality Matters in Understanding Hunger
The Hard Science Behind Our Satiety Algorithm
The Future of Satiety: What This Means for You
Learn More About High-Satiety Foods

The Ultimate Goal

The ultimate goal of our food addiction-satiety algorithm is to estimate how much you’ll eat based on what you eat. We use this understanding to optimise food and meal choices, boosting satiety and empowering our Optimisers to eat less without relying on unsustainable willpower.

And by golly, we’re getting close to the perfect algorithm that can forecast how much you’ll eat based on what you eat.

The chart below illustrates the relationship between the actual calories from our extensive dataset and the calories predicted by our updated algorithm. It displays 99.9th percentile error bars, but the error bands are barely visible due to the strong relationship, given the large amount of data.

For those who want to understand more about it, this article delves into how it works and how we developed it.

What Drives Your Hunger? The Key Factors of Satiety

Others in the past have tried to develop rules of thumb to help us avoid overeating using single components, like:

Energy density – eat fruit, vegetables and soup,
Fibre – high-fibre foods are hard to overeat,
Sugar – avoid sugar because it’s addictive,
Salt – limit sodium because it’s so seductive,
Carbs/fat – limit them, and you’ll eat less or
Protein leverage – focus on foods with a higher protein % for satiety.

Our food addiction–satiety algorithm utilises all these factors to create a perfectly balanced system that can be applied to any dietary approach.

The pie chart below shows the significance of each component in our system. Protein has the highest weighting, but fibre, sugar, calcium, potassium, sodium, vitamin B2, and energy density are statistically significant.

While the single-factor approaches are somewhat helpful, combining all the factors helps us more accurately understand how much we’ll eat based on what we eat.

The Science of Cravings: Understand Your Nutrient Bliss Point

The breakthrough in this update of our satiety model is the understanding that certain nutrients are not necessarily “good” or “bad”.

More vs less is not necessarily better or worse. Instead, we crave a unique ‘bliss point’ for each nutrient. This bliss point is the concentration of a nutrient that aligns with maximum energy intake.

If a food hits the bliss point for multiple nutrients, it stimulates more dopamine, which makes us want to eat more.

These bliss points foods would be perfect if you’re a starving caveman or a bear who needs to fatten up for winter, but today, we’re surrounded by them. And, as they say, resistance is useless.

We have modelled this bliss point/satiety precisely using a combined Gaussian + exponential decay function. The coefficients for each curve are tailored to maximise the fit to 872,925 days of data from free-living people eating a range of diets worldwide!

Sodium: The Hidden Driver of Cravings

Let’s first look at sodium as an example. Notice the bliss point peak at the perfect concentration of sodium that we crave.

Our sense of taste is well-calibrated to determine how much of each nutrient we need — not too little, not too much — and we crave the Goldilocks zone of each essential nutrient.

Foods without sodium can often taste bland, so we tend to eat less of them. However, if we accidentally add too much sodium (imagine accidentally spilling the whole salt shaker on your steak), the food will taste ‘too salty’.

As you can see in the bliss point/satiety chart above, the data shows that we crave, on average, 2.9g of sodium per 2000 calories. Anything with this concentration of sodium will maximise intake.

If you actively try to reduce your sodium intake, there is a risk that your cravings for salty foods will only intensify until you lose the willpower battle and end up face-down in salty foods like chips.

The secret to mastering satiety is to move beyond the bliss point concentration to eat less while still getting the nutrients required to satisfy your cravings, which is what the Optimal Nutrient Intake of 4 g/2000 calories is designed to do.

But it doesn’t stop there.

Sugar: The Sweet Spot That Keeps You Coming Back for More

Let’s look at sugar next.

We have a strong preference for sweet foods because they provide fast-acting energy. But, like salt, adding too much sugar to our tea makes it ‘too sweet’.

The data reveals a bliss point at around 22% of our daily energy intake. We’ll eat less if our daily food intake contains more or less than this bliss point concentration.

Protein: The Satiety Powerhouse That Keeps You Full

Professors Ruabenheimer and Simpson describe in their 2005 Protein Leverage Hypothesis paper that we have an intense craving for protein.

But it doesn’t take much to satisfy our minimum survival needs. Once we get around 12% of our daily energy needs from protein, we stop craving it and seek energy from fat and carbohydrates.

There is only so much lean steak or chicken breast you can eat. Once you hit the minimum amount of protein you need, you experience sensory-specific satiety. We eat a lot less if we only have protein-rich foods available!

Calcium: More Than Just Strong Bones—It Affects Your Hunger, Too!

After protein, calcium is the second most abundant stored nutrient (e.g. in our teeth and bones), so it makes sense that our body might regulate our intake of it.

The data indicate a bliss point for calcium at approximately 1200 mg per 2000 calories. Foods that contain more calcium per calorie are often more satisfying.

While dairy tends to dominate the high-calcium foods category, plant-based foods contain more calcium per calorie.

Fibre: The Natural Appetite Regulator

Fibre isn’t an essential nutrient, but high-fibre foods are hard to overeat. The bliss point for fibre seems to align with the perfect texture and mouthfeel of ultra-processed foods. Meanwhile, an extremely low-carb diet with no fibre tends to be less palatable.

Potassium: The Underrated Nutrient That Shapes Your Appetite

In 2022, Raubenheimer and Simpson stated that animals (including humans) possess specific appetites for protein, carbohydrates, fat, and at least two micronutrients — sodium and calcium. However, they also noted that specific appetites for other nutrients likely exist.

When I interviewed Professors Raubenheimer and Simpson, they mentioned that the body may seek to regulate potassium to balance sodium.

Potassium is the yin to sodium’s yang. However, we don’t crave it like sodium, potassium powder, and potassium-rich foods taste bitter. So, we seem to have a natural aversion to potassium-rich foods once we get enough.

We also don’t fortify our foods with potassium, so potassium is likely a signature of minimally processed, nutrient-dense whole foods. As shown in the chart below, the data reveal a bliss point for potassium of around 1800 mg per 2000 calories. To the right, we can see that we eat much less foods with a higher potassium concentration as sensory-specific satiety sets in.

Our Optimal Nutrient Intake for potassium (4.5 g/2000 calories) is designed to be a stretch target to boost foods that contain more potassium, an essential nutrient that many people don’t get enough of in our modern food system.

Vitamin B2: The Surprising Nutrient That Influences Satiety

You may be surprised to learn that vitamin B2 is a statistically significant satiety factor until you see the foods with a high concentration of it.

It’s not that adding a multivitamin with a ton of B2 to make your pee bright yellow will boost satiety if you’re eating nothing but McDonald’s. Still, the data indicate a bliss point for foods containing B2, which aligns with the Dietary Reference Intake. Moving towards the Optimal Nutrient Intake of 2.75 g of B2 per 2000 calories aligns with eating significantly less.

Energy Density: Why Some Foods Are Harder to Stop Eating

Interestingly, we also see a bliss point for energy density. Fruits and vegetables (located towards the right of the chart) contain a high amount of water and minimal calories, making them difficult to overeat.

Meanwhile, lower-carb, higher-protein foods, located towards the left of the chart, are more challenging to overeat.

Again, we observe an energy density bliss point that corresponds to maximum energy intake. Foods that hit this energy-density bliss point tend to contain a blend of carbs and fat.

What About the Rest? The Bigger Picture of Satiety

I think it’s likely that we may experience cravings for almost any nutrient if we are deficient in it. However, the nutrients identified by the analysis appear to be the signature of the foods and nutrients we crave the most. Conversely, they will satisfy those cravings with less energy.

It’s not that the other nutrients aren’t essential, but the shortlist of seven statistically significant nutrients elucidates the signature of higher-satiety foods. We’ve used these to create our Food Addiction Satiety Index, and in our 20/20 Macros course, we show Optimizers how to increase satiety per calorie. Learning how to optimise your nutrient density is super important, but it is a little more complex and the focus of our 20/20 Micros course.

Principal Component Analysis: Find the Signal in the Noise

When creating an algorithm to calculate satiety, the key is identifying the statistically significant factors that allow us to improve the system’s accuracy and leave the others.

Principal Component Analysis (PCA), which reduces data to its most useful variables, is a concept used in machine learning and artificial intelligence.

The chart below shows a simplified PCA that only considers calories, protein, fibre, carbs, and fat. Notice how protein percentage is diametrically opposed to calorie intake, meaning that if you want to reduce calories, you should prioritise higher protein percentage foods. Meanwhile, foods that contain more fat and carbohydrates also tend to have higher calorie content and less protein.

When we examine the data with more micronutrients included, we observe that many are clustered together in the same foods. However, one nutrient often dominates that cluster. We’ve eliminated the less significant factors to focus on the shortlist of nutrients that yield the biggest bang for your buck, creating an efficient algorithm that works with commonly available nutrient data.

Data Quality Matters in Understanding Hunger

It’s also worth noting that the nutritional parameters that are statistically significant in our analysis tend to be the ones that the FDA mandates food manufacturers report on their packaging. Hence, the data quality of these nutrients is likely higher, impacting the statistical significance.

Previous research has established that we have a specific appetite for protein, sugar, sodium, and calcium, while potassium and vitamin B2 are less clear.

Perhaps in the future, we will have more complete data that will enable us to confirm that we indeed have a specific appetite and satiety response for many more nutrients, especially when our intake is significantly above or below the bliss point.

But until then, we should view this combined satiety algorithm as the quantitative signature of eating patterns that align with eating less rather than eating more.

Rather than singular nutrients, the statistically significant nutrients in the algorithm likely represent clusters of nutrients that define groups of foods and eating patterns that align with eating more rather than eating less. Importantly, we can’t simply hack the algorithm with supplements. We need to obtain more of these key nutrients, along with the others that come with them, from our food.

This approach is potentially revolutionary in nutrition because it enables us to shift from a focus on avoiding ‘bad things’ in our diet to emphasising the nutrients we need to thrive.

The Hard Science Behind Our Satiety Algorithm

Many people use the phrase, “correlation does not equal causation”; as you’ll see, this statistical analysis is incredibly robust.

As mentioned above, each satiety curve has been optimised to fit the 872,925 days of data, helping us understand how much energy we might consume based on a single variable. We’ve combined these with the optimal weighting to empower you to understand how much you’re likely to eat based on your current diet.

Summary of Statistical Analysis

The table below summarises the statistical analysis from the design of our satiety algo.

parameter	p-value	correlation	power	t-stat
combined	0	0.37	1.53	369
protein	0	0.29	0.69	284
potassium	0	0.21	0.29	203
riboflavin (B2)	0	0.18	0.28	175
calcium	0	0.17	0.40	164
sugar	0	0.16	0.50	153
fibre	0	0.14	0.25	135
energy density	0	0.13	0.22	122
sodium	0	0.07	0.21	62

p-value

The first thing to note is that the p-value for each curve is negligible. p < 0.005 is often used in smaller studies to validate that the relationship is not due to chance; however, with such a large dataset, we can be confident that this is not random.

Correlation Coefficient

The third column shows the correlation of each variable with actual calorie intake. Protein has the highest correlation with any single variable (0.29), but combining them all boosts the correlation to 0.37. This means we can explain a significant portion of the variance in how much you eat by knowing what you eat, regardless of factors such as size, muscle mass, gender, or activity levels.

Power Regression

In a recent email exchange, Professors Raubenheimer and Simpson suggested that I use a power regression to test the statistical significance of the relationship between each variable. The power regression transforms the variables onto a log-log scale. The closer this is to 1.0, the stronger the relationship.

The first thing to note here is that simply considering protein vs. energy gives us a power exponent of 0.21. However, once we implement the bliss point regression, this increases to 0.69. Hence, we can see that the Gaussian-exponential decay function vastly improves accuracy!

The table above shows that protein has the highest power exponent when considering each factor alone. Still, when we combine all of them, the power exponent increases to 1.53, indicating that the relationship between actual calories and predicted calories is highly significant.

t-stat

Finally, we have the t-stat, which compares the variance in two populations to assess if it’s due to chance. Typically, a value greater than 3 is considered robust evidence. The t-stat for sodium is the lowest at 62, but the combined system gives us a massive t-stat of 369, indicating that the analysis is incredibly robust!

The Future of Satiety: What This Means for You

For years, diet advice has been overly simplistic—”eat less sugar,” “eat more fibre,” or “just control your portions.” But our cravings aren’t random; they’re biologically programmed to seek out specific nutrients in precise amounts.

With 872,925 days of data, we’ve uncovered a fundamental truth: satiety isn’t just about avoiding certain foods—it’s about hitting the right balance. Our bliss point model reveals that certain nutrients influence appetite in predictable ways, enabling us to understand why we eat what we do.

Key Takeaways

Your cravings follow a pattern. The foods you crave most are those that hit multiple nutrient bliss points.
Satiety is a science, not a mystery. We now have a data-backed formula to predict how much you’ll eat based on nutrient composition.
Smarter food choices = effortless appetite control. When you move beyond the bliss point of certain nutrients, you naturally eat less without feeling deprived.

What’s Next?

As you can see, we’ve worked hard to create the most accurate satiety algorithm possible, utilising all available data and nutritional factors.

Food Lists and Infographics

But if you’re looking for guidance on what to eat, check out our simple high-satiety infographics and food lists here.

Popular Foods – Interactive Search Tool

For more details, check out our interactive search tool containing the 750 foods most popular with our Optimiesrs.

Complete USDA Database: 8535 Foods

If you want ALL the data, we just published an interactive chart showing where all 8,535 foods in the USDA database sit on the spectrum of satiety vs. nutrient density, which you can view here.

20/20 Macros

However, after years of working with Optimisers, we’ve found that the best way to help them increase satiety is to upgrade their core list of foods and meals they consume regularly.

In 20/20 Macros, we’ll show you how to overcome food addiction and crush your cravings by making incremental changes to your current eating habits to discover Your Optimal 20/20 foods and meals that you love eating every day! We’d love you to join us.