Food Additive Mixes and Diabetes Risk

Food additives help manufacturers preserve products, maintain texture, adjust flavor, improve appearance, and extend shelf life. Each additive is normally assessed individually before regulators authorize its use, but that is not always how people encounter these substances in everyday life.

A packaged dessert, sauce, flavored drink, or ready-made meal may contain several additives at once. Across an entire day, a person may consume multiple combinations repeatedly. This raises an important question: Could frequently consumed mixtures affect health differently from the individual ingredients considered separately?

A large study published in 2025 examined whether common combinations of food additives were associated with the development of type 2 diabetes. The results do not prove that these additives cause diabetes, but they highlight a gap in how researchers and regulators evaluate real-world food exposure.

Why Are Additives Used?

Food additives perform a wide range of technical functions. Preservatives can slow spoilage and reduce the risk of foodborne illness. Emulsifiers and stabilizers help ingredients remain evenly mixed. Acidity regulators alter flavor or improve stability, while colors and sweeteners influence appearance and taste.

Both natural and synthetic substances may be classified as additives. Pectin occurs naturally in fruit, curcumin is found in turmeric, and anthocyanins give many fruits and vegetables their color. Their presence on an ingredients list does not automatically make a food harmful—or healthy.

International and national authorities evaluate additives using toxicological, exposure, and other available safety data. They may establish an acceptable daily intake or restrict how much can be used in particular foods. World Health Organization

However, traditional assessments have generally focused on one substance at a time. They may provide less information about combinations that occur across processed foods or about metabolic outcomes arising after years of repeated exposure.

Yet people do not construct their diets one additive at a time. This is why researchers from the French NutriNet-Santé project decided to examine patterns of additives that were commonly consumed together.

How the Study Worked

The researchers analyzed information from 108,643 French adults enrolled in the NutriNet-Santé cohort between 2009 and 2023. The average participant was approximately 42 years old, and about 79% of the participants were women.

Participants regularly completed detailed 24-hour food records. Whenever possible, they reported the brand and commercial name of packaged foods. Researchers combined this information with food-composition databases and laboratory measurements to estimate exposure to specific additives.

The team focused on 75 additives consumed by at least 5% of the participants. A statistical technique was then used to identify five combinations that tended to appear together in people’s diets.

These were not five fixed recipes deliberately given to participants. They were data-derived patterns representing additives that people were likely to consume together because they appeared in the same products or in foods commonly eaten by the same individuals.

Participants were followed for an average of 7.7 years. During that period, 1,131 people were diagnosed with type 2 diabetes.

The analysis accounted for many factors that could influence diabetes risk, including age, education, physical activity, smoking, body size, family history, energy intake, sugar, saturated fat, alcohol, and overall nutritional quality.

Which Mixtures Were Associated With Diabetes?

Two of the five identified mixtures were associated with a higher incidence of type 2 diabetes. The other three showed no statistically significant association.

Mixture 2: Emulsifiers and Stabilizers

The second mixture was characterized primarily by:

• Modified starches
• Pectin
• Guar gum
• Carrageenan
• Polyphosphates
• Xanthan gum
• Potassium sorbate
• Curcumin

Many of these additives function as emulsifiers, thickeners, stabilizers, preservatives, or colors. This pattern was particularly associated with foods such as industrially produced broths, dairy desserts, fats, and sauces.

For every one-standard-deviation increase in the mixture score, the researchers observed an approximately 8% higher relative incidence of type 2 diabetes.

This does not mean that eating a product containing pectin or xanthan gum raises an individual’s diabetes risk by 8%. The figure describes a statistical change connected with the overall mixture score across the study population, not the effect of one serving or one ingredient.

Mixture 5: Additives Commonly Found in Drinks

The fifth mixture was strongly connected with both sugary and artificially sweetened beverages. It included several categories of additives.

Acidifiers and acidity regulators included:

• Citric acid
• Sodium citrates
• Phosphoric acid
• Malic acid

Artificial sweeteners included:

• Acesulfame potassium
• Aspartame
• Sucralose

Colors and coating agents included:

• Sulphite ammonia caramel
• Anthocyanins
• Paprika extract
• Carnauba wax

Emulsifiers and stabilizers included:

• Gum arabic
• Pectin
• Guar gum

A one-standard-deviation increase in this mixture score was associated with an approximately 13% higher relative incidence of type 2 diabetes.

Again, this was not a 13-percentage-point increase in absolute risk, and it does not demonstrate that every listed additive contributed equally. The mixture may also reflect dietary behaviors that are difficult to measure completely.

The full peer-reviewed study and its detailed statistical results are available in PLOS Medicine.

Could the Additives Be Interacting?

The researchers found that the associations remained after adjusting for sugar, saturated fat, energy, alcohol, and broader diet quality. They also reported that no single additive appeared to explain the full association.

Exploratory analyses identified possible interactions between certain additives. Some appeared synergistic, meaning their combined statistical association was stronger than expected. Others appeared antagonistic, meaning one might have weakened the association involving another.

These interaction results should be interpreted cautiously. The researchers tested many possible pairings, and relatively few reached their selected statistical threshold. Experimental research is needed before scientists can determine whether these interactions represent genuine biological effects.

One possible pathway involves the gut microbiome. Animal and laboratory studies suggest that certain emulsifiers or sweeteners may alter microbial communities, intestinal inflammation, or glucose metabolism. Guar gum and carrageenan, for example, have produced metabolic or inflammatory changes in some animal experiments.

But results from cells or animals cannot automatically be applied to humans. The new study did not directly measure whether microbiome changes explained the observed diabetes associations.

Association Is Not Causation

But even a large, carefully conducted study cannot eliminate every alternative explanation. This was an observational investigation, not a randomized trial. Researchers recorded what participants ate and watched what happened over time; they did not assign people to consume particular additive combinations.

People with greater exposure to these mixtures may have differed in other ways. Although the analysis adjusted for many known risk factors, unmeasured lifestyle, health, socioeconomic, or dietary differences may still have influenced the findings.

Food records are also imperfect. Participants may forget certain products or report amounts inaccurately. Additive exposure was estimated using product information, databases, and laboratory data rather than measured directly in blood or urine.

The cohort contained a high proportion of women and people who were particularly interested in health and nutrition. The results may therefore not transfer directly to every country or population.

Only the 75 most widely consumed additives were included in the mixture model. Less common additives and combinations could not be examined reliably. Some cases of undiagnosed diabetes may also have been missed.

Most importantly, this was the first major prospective study designed specifically to connect real-life additive mixtures with type 2 diabetes incidence. The findings need to be reproduced in other populations and investigated in controlled human and laboratory studies.

What Should Consumers Do?

So these results are not a reason to fear every unfamiliar ingredient or remove all additives from your diet. Some additives play valuable roles in food safety, and an ingredient’s chemical-sounding name says little about its health effects.

The findings do provide another reason to limit frequent reliance on heavily processed foods containing long combinations of nonessential additives. This is especially relevant to sugary and artificially sweetened soft drinks, packaged dairy desserts, instant broths, and highly formulated sauces.

Practical steps include:

• Making water or unsweetened drinks your usual choice.
• Choosing whole or minimally processed foods more often.
• Comparing similar products and selecting the option with a simpler ingredients list when practical.
• Eating more vegetables, whole fruit, legumes, nuts, seeds, and whole grains.
• Treating packaged desserts, sweet drinks, and highly processed snacks as occasional rather than default choices.
• Considering the complete nutritional value of a food instead of judging it by one additive.

Someone who occasionally consumes a product containing guar gum, pectin, or citric acid should not conclude that the food will cause diabetes. Overall dietary patterns, body composition, physical activity, sleep, genetics, age, and family history remain important influences on type 2 diabetes risk.

The Bottom Line

This large French study identified five common food-additive mixtures. Two—one dominated by emulsifiers and stabilizers and another associated mainly with sweetened drinks—were linked with higher type 2 diabetes incidence.

The research is important because it studies additives as people actually consume them: in combinations. However, it demonstrates association rather than causation and cannot establish that any individual additive is dangerous at permitted levels.

For now, the most reasonable response is not alarm but proportion. Favor minimally processed foods, reduce regular intake of heavily formulated products, and allow future research to determine whether additive interactions have a direct role in metabolic disease.