What if you could rank food by ‘healthiness’ as you shopped? Nutrient profiling systems use algorithms to simplify picking healthy groceries

Imagine a world where food on grocery store shelves is ranked by its healthiness, with simple, research-backed scores. In some countries, that world already exists.

Nutrient profiling systems, or NPSs, support clear front-of-package labels that assess food quality based on nutrient content. Nutri-Score in France is a rainbow-colored system grading foods from A to E. Health Star Rating in Australia is a five-star system rating foods in half-star increments. And the Traffic Light System in the U.K. labels nutrient levels as green, yellow or red.

- Advertisement -

In contrast, the U.S. lacks a front-of-package ranking system for food. Food Compass was recently developed out of Tufts University to help address this gap and shortcomings in other systems. But it uses nutritional information not currently available for most foods and consumers.

As a gastroenterologist and physician-scientist, I focus on making the latest microbiome and nutrition data more accessible to the public. Drawing on this research, I developed Nutrient Consume Score, or NCS, which rates foods from 1 to 100 using nutritional information available for all foods and incorporates factors important for a healthy microbiome.

But how do nutrient profiling systems work? And how do they compare to other nutrition guides for consumers?

Nutrient cyphers

Each nutrient profiling system uses different scoring algorithms, but most assign positive points to nutrients and foods that are typically underconsumed, such as fiber, fruits and vegetables. Conversely, negative points are given to overconsumed nutrients like sugar, saturated fat and sodium, which are often added to processed foods. These points are combined into a single score: higher scores indicate healthier foods, while lower scores indicate less healthy options.

For example, kale – rich in fiber, potassium and unsaturated fats, while low in sugar, sodium and saturated fats – would earn a high score. In contrast, Twinkies, which are high in sugar, sodium and saturated fats, but low in fiber, potassium and unsaturated fats, would receive a low score. A food like black olives, high in fiber but also high in sodium, would fall somewhere in between.

Nutrient profiling systems work similarly to the Nutrition Facts labels on the back or sides of food packages in helping consumers make informed choices. These labels provide information about a food’s nutrient content, including calories, macronutrients, and key vitamins and minerals. The values are determined through laboratory analysis and nutrient databases based on standardized serving sizes regulated by the Food and Drug Administration.

But NPSs differ in that they combine nutrition information into a single actionable score. This means you don’t have to spend time deciphering Nutrition Facts labels, which are often in small print and can be confusing to interpret.

Ultraprocessed profiling

Nutrient profiling system algorithms are all quite similar in their high ranking of unprocessed foods – beans, nuts, seeds, fruits, vegetables and whole grains – and low ranking of processed foods like hot dogs, soft drinks, cakes and cookies. They help people rebalance their diets that have been skewed by food processing, or the degree to which the ingredients have been altered.

They complement the NOVA classification system developed by researchers at the University of São Paulo, which categorizes foods based on their level of processing. This system introduced the term “ultra-processed foods,” which are foods that have undergone significant industrial processing and contain ingredients not typically found in home cooking.

While NOVA has linked ultra-processed foods to poor health outcomes like obesity, worse mental health, cancer and early death, it treats all such foods equally, overlooking differences like amount of sugar, sodium and other additives.

Nutrient profile systems help provide nuance by identifying healthier options within the ultra-processed category. For example, plant-based milks, such as almond or soy milk, may be classified as ultra-processed under the NOVA system, but they can have relatively higher NPS scores if they contain minimal added sugars and salt.

Ratios and bioactives in balance

While nutrient profiling systems can be useful for choosing healthier options, current systems have limitations. They don’t always align perfectly with other research, often overlook the bioactive chemicals that regulate microbiome and body processes, and may rely on incomplete data. Current systems also don’t account for the caloric and health effects of alcohol.

The Nutrient Consume Score I designed aims to address these gaps by incorporating these neglected components of food. For example, it uses food categories as proxies for areas with limited data, including bioactive compounds like polyphenols, omega-3 fats and fermentable fibers. Proxies for bioactive compounds found in unprocessed foods – such as fruits, vegetables, grains, beans, nuts and seeds – are integrated into the score’s core algorithm, which uses nutrient ratios to measure the degree of food processing.

Nutrient ratios – including carbohydrate-to-fiber, saturated fat-to-unsaturated fat and sodium-to-potassium – reflect the natural balance of nutritional content of the cells in unprocessed foods, which research has shown correlate with cardiometabolic health.

For example, the cell walls of plants provide structural strength and are rich in fiber, while their energy vesicles store carbohydrates. Fiber reduces sugar absorption and is fermented into the compound butyrate, which maintains blood sugar and regulates appetite.

The fat profiles of unprocessed foods are similar to the fat composition in cell membranes. Saturated fat-to-unsaturated fat ratios capture how different types of fat, affect inflammation and weight.

Finally, the potassium-to-sodium ratio reflects the natural function of cell membrane pumps, which concentrate potassium inside cells while transporting sodium out. This affects blood pressure as well as microbiome and metabolic health.

Research currently under peer review shows that the Nutrient Consume Score compares favorably with other systems. Derived from nutrition data from nearly 5,000 Americans, NCS sores are linked to blood pressure, waist circumference and weight. NCS has also been incorporated into a smartphone app intended for public use, currently in beta testing.

Empowering smart choices

While nutrient profiling systems are a promising tool for healthier food choices, they come with important caveats. Most studies testing how well they work focus on how two factors relate to each other rather than whether one directly causes the other. Correlation doesn’t prove causation.

Further studies are needed to assess whether these systems influence buying habits, consumption trends, and health outcomes like weight and blood pressure. Additionally, individual dietary needs can vary, and personalized algorithms could help refine these scores for tailored recommendations.

Despite these considerations, nutrient profiling systems are promising tools to combat rising rates of metabolic disease. Their use in Europe demonstrates their potential to shift consumer purchasing habits and inspire food companies to create healthier products.

Americans may one day see similar front-of-package labels in the U.S. Until then, smartphone technologies can offer a practical way to help consumers make smarter choices today.