Why Is Caramel Color Bad for You?

Caramel color itself is not a single substance, and the health concerns are not about the brown tint it imparts to your food. The worry centers on chemical byproducts that form during the manufacture of certain types of caramel color, particularly a compound called 4-methylimidazole (4-MEI), which the International Agency for Research on Cancer has classified as “possibly carcinogenic to humans.” A second contaminant, THI, has been shown to suppress immune function in animal studies. Whether these byproducts pose a real threat at the levels you actually consume is where the story gets genuinely complicated.

Four Classes of Caramel Color, and Only Some Are Controversial

Caramel color is the world’s most widely used food coloring, and it comes in four distinct classes depending on how it is manufactured. Class I (also called plain caramel or E 150a) is produced simply by heating sugars without any reactive chemicals. Class II uses sulfite compounds during production. Class III is made with ammonia. Class IV, the type found in most colas and dark sauces, is made with both ammonia and sulfite compounds. The international food safety body JECFA considers Class I safe enough that it does not even set a numerical limit on daily intake, while Classes II through IV have specific acceptable daily intake ceilings.¹

The controversy almost entirely involves Classes III and IV. When sugars react with ammonia at high temperatures, they trigger a cascade of chemical reactions that generate 4-MEI as a byproduct.² Class III production also generates THI, the immune-suppressing contaminant. So when people ask whether caramel color is bad for you, the accurate answer depends on which class is in the product you are eating or drinking. A caramel-colored craft beer using Class I coloring and a can of cola using Class IV are not the same conversation.

The Cancer Question Around 4-MEI

The cancer concern traces back to a long-term feeding study conducted by the U.S. National Toxicology Program. Researchers fed 4-MEI to rats and mice at high dietary concentrations over two years. Male and female mice developed lung tumors at significantly increased rates, which the NTP described as “clear evidence of carcinogenic activity.” Rats, however, showed no clear cancer signal, with only a weak, ambiguous hint of leukemia in female rats.³ Based on the mouse findings, IARC classified 4-MEI as a Group 2B carcinogen in 2011, meaning “possibly carcinogenic to humans.”⁴

“Possibly carcinogenic” sounds alarming, but Group 2B is IARC’s lowest positive classification. It means there is limited evidence from animal studies and insufficient evidence from human studies. The doses that caused lung tumors in mice were vastly higher than what people actually consume through food and beverages.⁵ That dose gap is the core of the regulatory debate: the effect is real in mice at extreme exposures, but whether it translates to any meaningful risk at human dietary levels remains unresolved.

Why Mice Get Lung Tumors and What That Means for People

One reason the 4-MEI cancer question remains unsettled is a stark species difference. Mice developed lung tumors, but rats fed the same compound at comparable doses did not. This is not unusual in toxicology. Mice are known to be more susceptible to certain types of lung tumors than rats or humans, and some researchers have argued that the mouse lung tumor response reflects a species-specific vulnerability rather than a universal carcinogenic mechanism.⁵

A comprehensive toxicological review noted that experimental models show 4-MEI can damage the liver through oxidative stress and inflammation, and that California classifies it as a possible carcinogen with strict exposure limits, while the European Food Safety Authority has found the genotoxic evidence insufficient to draw firm conclusions.⁶ In other words, different regulatory bodies looking at the same data have landed in genuinely different places, which is part of why consumers find the topic confusing.

DNA Damage at High Concentrations

Beyond the cancer studies, researchers have looked at whether 4-MEI can directly damage DNA, a finding that would strengthen the case for it being a true carcinogen rather than a compound that simply promotes tumor growth through other means. The picture here is messy. Standard mutagenicity assays, the workhorse tests for detecting whether a substance causes genetic mutations, have largely come back negative.⁶

However, when researchers exposed human white blood cells to 4-MEI in a lab dish, they found evidence of chromosomal damage and other DNA abnormalities at the higher concentration levels tested. At the very highest concentration, 4-MEI actually caused more of one type of DNA damage than the positive control chemical used as a benchmark.⁷ The catch is that these were in-vitro tests at concentrations far above what would circulate in your bloodstream from drinking cola. The genotoxicity findings add a note of caution but fall short of proof that dietary 4-MEI damages human DNA in real-world conditions.

THI and Your Immune System

4-MEI gets most of the headlines, but a second contaminant deserves attention. THI (2-acetyl-4-tetrahydroxybutylimidazole) is found specifically in Class III caramel color, the type made with ammonia but without sulfites. Animal feeding studies dating back decades have shown that THI causes a drop in white blood cell counts, particularly lymphocytes, which are central to your immune defense.⁸

The mechanism has since been worked out. THI inhibits an enzyme called S1P lyase, which plays a role in regulating how lymphocytes move through and out of lymph nodes. Blocking that enzyme traps lymphocytes where they are, depleting them from the bloodstream.⁹ This is a well-characterized immunosuppressive effect, not a vague association. The good news is that in rat studies, white blood cell counts returned to normal after exposure stopped, suggesting the effect is reversible.⁸

How much THI are people actually consuming? A dietary intake assessment of the Chinese population estimated average THI exposure from Class III caramel colors at a fraction of a microgram per kilogram of body weight per day, and concluded that the levels were of low health concern based on existing toxicological data.¹⁰ That said, this is a population average. People who consume very large amounts of products colored with Class III caramel could theoretically push into higher exposure territory, though we lack human studies showing immune suppression at any realistic dietary intake.

Where Your Exposure Actually Comes From

If you are thinking “this is mostly about cola,” you are partly right but missing the bigger picture. A U.S. dietary exposure assessment found that cola-type carbonated beverages were the largest single source of 4-MEI from added caramel color for most population groups. But 4-MEI also shows up in foods you would not immediately associate with food coloring. Coffee was the top contributor among foods where 4-MEI forms naturally during cooking or roasting rather than being added as an ingredient.¹¹

An updated assessment broadened the contributor list further. When researchers adjusted 4-MEI levels beyond just beverages and looked at labeled food products more broadly, the top three contributors to added 4-MEI exposure were prepared meals, breads and rolls, and ice cream and frozen yogurt.¹² Caramel color is added to a surprisingly wide range of products to achieve a consistent brown appearance, from soy sauce and gravies to baked goods and pet food. And any food that undergoes high-heat browning, like bread crusts and roasted coffee, generates some 4-MEI on its own through naturally occurring chemical reactions between sugars and amino acids.

This dual-source problem matters because it means you cannot eliminate 4-MEI exposure entirely, even if you avoid every product listing caramel color on the label. The thermal treatment pathway means that any food subjected to intense browning produces trace amounts.

What Regulators Have Actually Done

Regulatory responses to caramel color contaminants vary considerably depending on where you live. The most aggressive action has come from California, where Proposition 65 requires beverages to carry a cancer warning label if they contain 4-MEI at levels that correspond to an excess cancer risk greater than one case per 100,000 exposed people, which works out to a daily threshold of 29 micrograms.¹³ That regulation pushed several major soda manufacturers to reformulate their caramel color processes to lower 4-MEI levels, and the same study found that these state regulatory standards appeared to be effective in reducing exposure.

The FDA, by contrast, has not set a specific limit on 4-MEI in food. The agency’s position has been that the levels of 4-MEI in foods are not high enough to pose a health risk, though it has continued monitoring the science. The European Food Safety Authority re-evaluated all four classes of caramel color and set acceptable daily intake limits, but did not ban any class. JECFA’s acceptable daily intake for Classes III and IV stands at up to 200 milligrams of the caramel color itself per kilogram of body weight per day.¹ These are limits on the total caramel color, not on the 4-MEI contaminant within it, which adds another layer of complexity.

The regulatory disagreement is genuine and reflects real scientific uncertainty. California uses a precautionary approach, essentially saying “we are not sure this is safe, so we will set tight limits.” Most other regulators take a risk-based approach, saying “the animal evidence requires extremely high doses that consumers do not approach.” Neither position is irrational given the available data.

Effects on the Gut

A newer area of research looks at how caramel color interacts with the communities of bacteria in your digestive system. A review of food additives in ultra-processed products noted that caramel color IV and its byproducts have been associated with changes in the composition and abundance of gut microbiota.¹⁴ This is a preliminary finding, and the evidence is thin compared to the cancer and immune data, but it aligns with a broader pattern of concern about how synthetic food additives affect gut health.

Interestingly, not all caramel-related compounds appear harmful to the gut. Research on a specially enriched caramel containing high levels of a sugar derivative called di-D-fructose dianhydride found that it actually promoted beneficial gut bacteria, including lactobacilli and bifidobacteria, in rats with induced colitis. The enriched caramel also reduced markers of inflammation and tissue damage.¹⁵ This was a highly specific experimental caramel, not the kind used in commercial food products, so it would be wrong to conclude that caramel color is good for your gut. But it does illustrate that the chemistry of caramelized sugars is complex, and the effects depend heavily on the specific compounds present.

Can Manufacturers Reduce the Problem?

One underappreciated aspect of this issue is that 4-MEI levels in caramel color are not fixed. They depend on manufacturing conditions, and there is active research into how to reduce them. For instance, a study on the chemistry of 4-MEI formation found that adding certain divalent metal ions during the caramel-making process could inhibit the chemical pathway that generates 4-MEI. The metals appeared to redirect the sugar-ammonia reaction toward forming melanoidins (the brown pigments manufacturers actually want) rather than breaking down into the small molecules that include 4-MEI.¹⁶

The California Proposition 65 experience demonstrated that regulatory pressure can drive reformulation. After the labeling requirement took effect, several beverage companies modified their caramel color production processes to bring 4-MEI levels below the warning threshold.¹³ This suggests that the amount of 4-MEI in a given product is a choice, not an inevitability. A beverage company selling cola in California and the same cola in another state could theoretically be using different caramel color specifications, though many companies now use the lower-4-MEI formulations everywhere for simplicity.

Practical Takeaways for Everyday Choices

If you drink an occasional cola or use soy sauce on your stir-fry, the science does not suggest you need to worry. The exposure levels associated with harm in animal studies are orders of magnitude higher than what a typical diet produces. Even high-consumption estimates for the general population fall well below the thresholds regulators use to flag concern.¹⁰

That said, if you want to minimize your exposure, the most effective step is reducing your intake of cola-type soft drinks, since these are the single largest added source of 4-MEI for most people. Choosing lighter-colored beverages, checking ingredient labels for “caramel color” in processed foods, and being aware that dark-colored sauces, gravies, and baked goods often contain it are all reasonable strategies. You cannot avoid 4-MEI entirely because it also forms naturally during cooking and roasting, but the naturally formed amounts tend to be smaller than what gets added through commercial caramel coloring.

The broader context matters here too. Caramel color is used in enormous volumes globally, so even a small per-serving risk, if it exists, could affect a large number of people. This is why advocacy groups push for stricter limits and why California’s approach is more precautionary than most. Whether you find that level of caution appropriate depends partly on your personal risk tolerance and partly on how much weight you give animal studies at high doses as predictors of human outcomes at low doses. The honest scientific answer is that caramel color contaminants are not proven to harm humans at dietary levels, but the animal data raises enough questions that the topic deserves continued scrutiny rather than dismissal.

1
Europe PMC / Springer (Journal of Food Science and Technology). Food caramels: a review
2
PubMed. Carcinogenic 4(5)-methylimidazole found in beverages, sauces, and caramel colors: chemical properties, analysis, and biological activities
3
PubMed. Toxicology and carcinogenesis studies of 4-methylimidazole (Cas No. 822-36-6) in F344/N rats and B6C3F1 mice (feed studies)
4
PubMed. A weight of evidence assessment of the genotoxic potential of 4-methylimidazole as a possible mode of action for the formation of lung tumors in exposed mice
5
PubMed. Evaluation of the mode of action of mouse lung tumors induced by 4-methylimidazole
6
PubMed. Toxicological evaluation of 4-methylimidazole: Research advances, health concerns and regulatory perspectives
7
PubMed. Genotoxic effects of 4-methylimidazole on human peripheral lymphocytes in vitro
8
Food and Chemical Toxicology. Toxicity studies of Caramel Colour III and 2-acetyl-4(5)-tetrahydroxybutylimidazole in F344 rats
9
PubMed. Component of Caramel Food Coloring, THI, Causes Lymphopenia Indirectly via a Key Metabolic Intermediate
10
PubMed. Dietary intake assessment of caramel colours and their processing by-products 4-methylimidazole and 2-acetyl-4-tetrahydroxy-butylimidazole for the Chinese population
11
PubMed. A U.S. population dietary exposure assessment for 4-methylimidazole (4-MEI) from foods containing caramel colour and from formation of 4-MEI through the thermal treatment of food
12
PubMed. An updated U.S. population dietary exposure assessment for 4-methylimidazole (4-MEI) from foods containing colour and from formation of 4-MEI through thermal treatment of foods
13
PubMed Central. Caramel color in soft drinks and exposure to 4-methylimidazole: a quantitative risk assessment
14
Revista médica del Hospital General de México. Food additives in ultra-processed products and some health effects
15
Journal of Agricultural and Food Chemistry. Di-d-fructose Dianhydride-Enriched Caramels: Effect on Colon Microbiota, Inflammation, and Tissue Damage in Trinitrobenzenesulfonic Acid-Induced Colitic Rats
16
PubMed. Effects of divalent cations on the formation of 4(5)-methylimidazole in fructose/ammonium hydroxide caramel model reaction