Emulsifiers, other additives and type 2 diabetes risk
What it found
In one large French cohort, some additive mixtures, emulsifiers, preservatives, colourings and sweeteners were linked to more new cases of type 2 diabetes.
Short human trials found some gut and blood marker changes, but no clear harm to blood sugar control.
What they found
Studies in people
Additive mixtures and diabetes
In 108,643 French adults, two of five additive mixture patterns were linked to more new cases of type 2 diabetes. One was dominated by emulsifiers, the preservative potassium sorbate and the colourant curcumin; the other by acidifiers, artificial sweeteners, colourants and some emulsifiers.
Emulsifiers in a cohort
In 104,139 French adults with 1056 new cases of type 2 diabetes, higher intakes of total carrageenans, guar gum, tripotassium phosphate, sodium citrate and xanthan gum were linked to new cases. The links were generally modest and differed between emulsifier types.
Preservatives and colourings
Among 17 preservatives studied in 108,723 French adults, 12 were linked to more new cases of type 2 diabetes, with the strongest links for potassium sorbate and sodium nitrite. Total colouring additives and several individual colourants were also linked to more cases.
Sweeteners in a cohort
Higher total artificial sweetener intake, and higher intakes of aspartame, acesulfame-K and sucralose, were each linked to new cases of type 2 diabetes in the same French cohort. The estimate for sucralose was smaller than the others.
Short human feeding trials
In 16 healthy adults given 15 g of carboxymethylcellulose daily for 11 days, gut bacteria changed but blood sugar control did not get worse. In 20 men given carrageenan for 2 weeks, the whole group showed no change in insulin sensitivity, though overweight men showed lower insulin sensitivity and more inflammation.
Sweetener trial in people
In 120 healthy adults given sweeteners for 2 weeks, saccharin and sucralose worsened blood sugar responses, while aspartame and stevia caused no significant change at group level. Responses varied a lot between individuals.
Animal and lab studies
Animal and lab work
Mice fed carboxymethylcellulose or polysorbate 80 for 12 weeks developed gut barrier changes, inflammation and features of metabolic syndrome, and these effects transferred to germ-free mice through faecal transplants. Lab tests on human gut bacteria found carrageenan and some other emulsifiers reduced bacterial diversity and raised inflammatory markers, while lecithin did not.
What the authors conclude
“Current evidence therefore supports biological plausibility, not causality or additive-specific clinical recommendations.”
Also in their conclusions
- They say the current evidence supports biological plausibility, not causality or additive-specific clinical recommendations.
- They call for longer controlled feeding trials that test realistic mixtures while holding the food matrix constant.
- They advise these trials to include exposure validation, repeated microbiome and metabolomic sampling, intestinal permeability measures and validated insulin-sensitivity endpoints.
How it was done
Food additives are eaten together in ultra-processed foods, and it is hard to tell their own contribution to blood sugar problems apart from overall diet quality and the food they come in. The authors wanted to weigh up the evidence from population studies, human trials and lab work.
They gathered and interpreted studies on food additives and type 2 diabetes, including population studies, human feeding trials, animal work and lab tests. The main population evidence came from NutriNet-Santé analyses, including one with 108,643 French adults followed for an average of 7.7 years, in which 1131 new cases of type 2 diabetes were found.
What it can’t tell you
- The population studies followed people over time, so they can show a link but not that the additives caused the diabetes.
- The human trials were short and small, so they cannot show whether any effects last or matter for health.
- The animal and lab results cannot show what happens in people eating everyday diets.
Who paid
- Conflicts
- The authors declare no conflicts of interest.
- Authors work at
- University of Split School of Medicine, Croatia
The paper
- Title
- Food Additive Mixtures, Glucose Metabolism, and Type 2 Diabetes Mellitus Risk: Mechanistic Insights from Epidemiology, Human Intervention Studies, and Experimental Models
- Type
- ReviewThe authors read earlier studies and describe what they found. It isn’t a new study.
- Evidence
- Studies in people · animals · lab
- Summarised from
- Full text
- Licence
- CC BY. Words in quotation marks are the authors’; the rest is our summary.
- Cite
- Santic R, Kumric M, Pavlovic N, Bozic J (2026). Food Additive Mixtures, Glucose Metabolism, and Type 2 Diabetes Mellitus Risk: Mechanistic Insights from Epidemiology, Human Intervention Studies, and Experimental Models. Nutrients. doi:10.3390/nu18152521Free full textPubMed 42588144DOI
Summary written 30 Sep 2026. Check it against the paper before it changes what you eat. How we summarise papers · Report an error