United States Food and Drug Administration Bans FD&C Red No. 3 (Erythrosine)
03 May 2025
On 15 January 2025, in response to a court order, the United States (U.S.) Food and Drug Administration (FDA) revoked the authorization for the use of FD&C Red No. 3 (erythrosine), giving food and pharmaceutical manufacturers until 15 January 2027 or 18 January 2028, respectively, to reformulate their products.
An overview of the story and scientific data behind this situation is instructive as companies look at the data for their currently approved color additives and new color additives in development. The value of an experienced understanding of regulatory compliance and risk assessment is underscored in helping to bring new color additives to market safely and successfully.
Background and Perspective
The FDA revoked the use of erythrosine on the basis of the Delaney Clause, which was enacted in 1958 as part of the Food Additives Amendment of 1958 to the Food, Drug, and Cosmetics Act (FD&C Act). This clause prohibits FDA authorization of a food additive that has been found to induce cancer in humans or animals. It simply states:
No additive shall be deemed to be safe if it is found to induce cancer when ingested by man or animal, or if it is found, after tests which are appropriate for the evaluation of the safety of food additives, to induce cancer in man or animal (FDA, 2025a)
In addition to the Delaney Clause, at the urging of the FDA, a similar clause was included in the 1960 system for the premarket review of additives used specifically for coloring foods, drugs, and cosmetics. The meaning of this color additive–specific clause was the same as the Delaney Clause, and states that any ingested color additive shall not be approved if it is “found by the Secretary to induce cancer when ingested by man or animal” (Merrill, 1997).
The Delaney Clause and the color additive–specific clause of 1960 were based on the understanding of carcinogenesis at the time of their enactment. The binary classification of compounds as either carcinogenic or non-carcinogenic by these clauses implies the now-outdated assumption that all carcinogens act by a genotoxic mechanism for which there is no generally acceptable level of exposure. These clauses are also based on the assumption that effects observed in laboratory animals are directly relevant to human health (Merrill, 1997; Krishan et al., 2021). Notably, the Delaney Clause and the color additive–specific clause of 1960 apply to the banning of substances based on hazard (i.e., “does it have the potential to cause cancer”), without consideration for potency, exposure, or the risk of harm to humans—all of which are components of a proper assessment of risk (i.e., the likelihood of causing cancer under specific circumstances). As outlined below, assessments of safety are not as black-and-white as previously believed, and our understandings of risk and cancer have developed over time. Nonetheless, the ability of the courts to require the revocation of permission for the use of erythrosine gives us all pause for thought regarding color additives.
Erythrosine and the Current Ban
FD&C Red No. 3 (erythrosine) is a synthetic color additive. It is currently permitted as a color additive in foods and ingested pharmaceuticals in the European Union and Canada. In the U.S., it has been used in foods and ingested pharmaceuticals dating back to before 1969 (Health Canada, 2024; EMA, 2025; FDA, 2025b). The recent FDA revocation of permission for the use of erythrosine in foods and ingested pharmaceuticals is in reference to a color additive petition submitted to the FDA in 2022; the petition included data showing that high doses of erythrosine caused the development of tumors in male rats through a hormonal mechanism.
These findings were not new to FDA. For over 30 years, the FDA has been aware of the role of erythrosine in the development of thyroid tumors in male rats—and male rats only. Decades ago, the finding of thyroid tumor development in male rats was noted to lack relevance to humans, with the FDA recently noting that “the available data does not raise safety concerns for humans” (FDA, 2025b). The FDA went on to note that these effects were not reported in other animals or humans and that “there is no evidence showing FD&C Red No. 3 causes cancer in humans.” In past cases in which synthetic color and flavor additives have been found to be animal carcinogens but present a de minimis risk to humans (i.e., a less than 1 in 1,000,000 excess lifetime risk of cancer), the FDA has claimed “authority to ignore the language of the [Delaney Clause] when literal application would be pointless” (Merrill, 1997).
Despite these long-standing arguments, the court ruled that the FDA is bound by the Delaney Clause (Merrill, 1997; Felter et al., 2020) and, as such, the FDA is obligated to consider the use of erythrosine in foods and ingested pharmaceuticals to be unsafe (FDA, 2025b).
A deep dive is provided in the next 2 sections to provide a better understanding of the scientific arguments around erythrosine, the next sections answer the questions “What do the data show about erythrosine and tumor formation?” and “Does the scientific community (and the FDA) consider these tumors to be relevant to humans?”
Erythrosine – Toxicological Data
In the 1980s, safety studies of erythrosine were conducted, forming the foundation of the body of knowledge pertaining to the safety of this color additive. In a lifetime study in which male and female rats were exposed to erythrosine in utero and given diets containing up to 4% erythrosine (providing approximately 2,464 mg/kg body weight/day, or 107,000 times the estimated human exposure in the U.S.) for 30 months, an increased incidence of non-malignant thyroid follicular adenomas was reported in high‑dose males (Borzelleca et al., 1987; Capen, 1989 cited in Capen, 2001). No thyroid tumors were reported at lower doses or in female rats.
The results of further studies show that erythrosine does not directly induce these thyroid tumors but acts by an indirect, hormonal mode of action to which male rats are particularly sensitive. This mode of action is well documented and supported by a lack of accumulation of erythrosine in the thyroid, negative genotoxicity studies, a lack of neoplastic responses in mice and gerbils, a lack of response in rats at dietary concentrations of <1% erythrosine (Capen, 1989, 2001), and a lack of tumors in other organs (Capen and Martin, 1989; Capen, 1999 – Capen, 2001; Foster et al., 2021).
Indeed, additional mechanistic studies support a hormonally-mediated mechanism of carcinogenesis for erythrosine in male rats, the initiating event of which is the inhibition of an enzyme in the liver (5'‑monodeiodinase) responsible for the conversion of thyroxine (T4; a prohormone) to the active thyroid hormone triiodothyronine (T3) (Hoyer and Flaws, 2019). This reduction in circulating levels of T3, and reduction in T3-mediated negative feedback on the hypothalamus-pituitary-thyroid axis, results in a compensatory increase in the release of thyroid-stimulating hormone (TSH) by the pituitary gland, which stimulates the production and release of thyroid hormones. When this pathway persists over long periods of time, the ongoing stimulation of the thyroid follicular cells leads to proliferative changes, and, eventually, tumors (Capen, 2001; Foster et al., 2021).
An adverse outcome pathway for TSH-induced thyroid tumors in rats has been proposed by Foster et al. (2021) and is supported by years of studies of various compounds found to disrupt thyroid homeostasis in rats. The molecular initiating event specific to erythrosine is the inhibition of tissue deiodinases, which is followed by key events including (i) decreased conversion of T4 to T3; (ii) decreased circulating T3 levels; (iii) increased circulating TSH; (iv) thyroid follicular hypertrophy; and (v) thyroid neoplasia. The progression through these events in male rats requires ongoing high-dose exposure to the compound causing the molecular initiating event.
Relevance of Thyroid Effects in Rats to Humans
The increased sensitivity of rats—males especially—to thyroid tumors, in comparison to humans, is influenced by important differences in the structure and function of the thyroid gland, as well as the handling of thyroid hormones.
Compared to female rats and humans, male rats have taller follicular epithelial cells, with smaller lumens containing less colloid (a stored reserve of thyroid hormones); these features indicate that male rats have more active thyroid glands with lower T4 reserves than humans and female rats (Döhler et al., 1979 – Capen, 2001; Foster et al., 2021). Circulating T4 has a shorter plasma half-life in rodents (12 to 24 hours) compared to humans (5 to 9 days), which has been attributed primarily to differences in transport proteins (Döhler et al., 1979 – Capen, 2001; Foster et al., 2021). In humans, over 99% of circulating T3 and T4 is bound to thyroid-binding globulin (TBG), resulting in a lower proportion of circulating T4 being unbound and active compared to rats (Foster et al., 2021). TBG is a high-affinity protein that is not present in rodents (circulating T3 and T4 bind to lower-affinity albumin and transerythrin in rodents) (Foster et al., 2021).
The differences in thyroid gland activity, T4 reserves, and plasma T4 half-lives contribute to the higher susceptibility of rats to disruptions in thyroid hormone homeostasis, which can result in lower T4 availability and compensatory increases in serum TSH. Such increases in TSH levels have been observed in male rats at doses of various thyroid-disrupting compounds >100-fold lower than those required to elicit an increase in serum TSH in other species (Lewandowski et al., 2004). However, even in the absence of thyroid-disrupting compounds, baseline serum concentrations of TSH in male rats are approximately 25 times higher than those in female rats and up to 120 times higher than those in humans (EPA, 2005; Foster et al., 2021). With their faster thyroid hormone turnover and lower reserves, these high TSH concentrations enable rats to maintain a higher level of thyroid hormone production to keep up with systemic demand (EPA, 2005; Foster et al., 2021).
The increased susceptibility of male rats to thyroid tumors induced by ongoing TSH stimulation is not specific to erythrosine and has been observed under other conditions of primary excessive TSH secretion or induced goiter (thyroid enlargement) (Axelrad and Leblond, 1955; Ohshima and Ward, 1984, 1986 – Capen, 2001). Interestingly, compared to rodents, healthy humans are relatively resistant to this type of TSH‑induced thyroid cancer, with conditions of increased serum TSH over long periods of time (e.g., iodine deficiency), in the absence of underlying thyroid conditions, resulting in goiter rather than thyroid cancer (Doniach, 1970; McClain, 1989; Curran and DeGroot, 1991 – Capen, 2001; Foster et al., 2021).
Ultimately, the evidence resulting from decades of study of the effects of erythrosine on male rat thyroid glands demonstrates that the observed tumors are hormonally-mediated and are specific to male rats, and are not relevant to human safety (FDA, 2025a).
The 2025 Court Ruling: What Does This Mean?
Despite the wealth of data indicating that thyroid tumors caused by long-term, high-dose exposure to erythrosine in male rats likely are of no relevance to humans, which was previously communicated by the FDA (FDA, 2025b), the recent court decision presents a challenge to the use of scientific information to come to food safety decisions. Thus, as a result of the enforcement of the Delaney Clause, this court decision signals that approved color additives could be coming under similar scrutiny and companies may therefore need to consider alternative ingredients for coloring purposes.
How Can Intertek Help?
Any color additive in food is deemed unsafe unless its use is either permitted by—or exempt from— regulation. With the FDA having to revoke the authorization for the use of FD&C Red No. 3 (erythrosine), food and pharmaceutical manufacturers now have until 15 January 2027, or 18 January 2028, respectively, to reformulate their products. After these dates, foods and pharmaceuticals containing erythrosine will be considered to be adulterated under the FD&C Act (FDA, 2025c,d).
Any replacement color additives not currently permitted for their intended use(s) will require premarket evaluation and approval by the FDA prior to use. Companies looking to go through the approval process will need to look intently at the safety information for their color additives and will need to provide a scientifically justified and well-positioned dossier for regulatory review. Intertek has extensive experience with risk assessments in support of color additives and would be happy to assist interested parties in preparing and submitting color additive petitions.
References
Axelrad AA, Leblonde CP (1955). Induction of thyroid tumors in rats by a low iodine diet. Cancer 8:339-367.
Borzelleca JF, Capen CC, Hallagan JB (1987). Lifetime toxicity/carcinogenicity study of FD & C red no. 3 (erythrosine) in rats. Food Chem Toxicol 25:723-733.
Capen CC (1989). Mechanistic considerations for thyroid gland neoplasia with FD&C Red No. 3 (erythrosine). In: 1989 European Meeting of the Toxicology Forum. Washington (DC): Toxicology Forum, Inc., pp. 113-130. Cited In: Capen, 2001.
Capen CC (1999). Thyroid and parathyroid toxicology. In: Harvey PW, Rush K, Cockburn A, eds. Endocrine and Hormonal Toxicology. Sussex: Wiley, 1999, pp. 33-66. Cited In: Capen, 2001.
Capen CC (2001). Toxic responses of the endocrine system (ch21). In: Klaassen CD, ed. Casarett and Doull's Toxicology: The Basic Science of Poisons, 6th ed. New York, Toronto: McGraw-Hill, Medical Publishing Division, pp. 711-759.
Capen CC, Martin SL (1989). The effects of xenobiotics on the structure and function of thyroid follicular and C-cells. Toxicol Pathol 17:266-293.
Curran PG, DeGroot LJ (1991). The effect of hepatic enzyme-inducing drugs on thyroid hormones and the thyroid gland. Endocr Rev 12:135-150.
Döhler KD, Wong CC, von zur Mühlen A (1979). The rat as model for the study of drug effects on thyroid function: consideration of methodological problems. Pharmacol Ther B 5:305-318. Cited In: Capen, 2001.
Doniach I (1970). Aetiological consideration of thyroid carcinoma. In: Smithers D, ed. Tumors of the Thyroid Gland. (Neoplastic Disease at Various Sites, Vol. VI). London: E & S Livingstone, pp. 55-72.
EMA (2025). List of substances [search results for: ‘erythrosine’]. In: IRIS: A Secure Online Platform for Handling Product-related Scientific and Regulatory Procedures with EMA]. Available at: https://iris.ema.europa.eu/substances/ [last accessed: March 3, 2025].
U.S. FDA (2025a). Federal Food, Drug, and Cosmetic Act (FD&C Act): Chapter 9. Subchapter IV - Food. 21 USC §348 - Unsafe food additives [Sec. 409]. In: U.S. Code-Title 21-Food and Drug (Food and Drug Administration). Washington (DC): U.S. House of Representatives, Office of Law Revision Counsel. Available at: https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title21-section348&num=0&edition=prelim.
FDA (2025b). FD&C Red No. 3 [Regulatory Information Available: https://www.fda.gov/industry/color-additives/fdc-red-no-3.
Felter SP, Llewelyn C, Navarro L, et al. (2020). How the 62-year old Delaney Clause continues to thwart science: Case study of the flavor substance β-myrcene. Regul Toxicol Pharmacol 115:104708.
Foster JR, Tinwell H, Melching-Kollmuss S (2021). A review of species differences in the control of, and response to, chemical-induced thyroid hormone perturbations leading to thyroid cancer. Arch Toxicol 95:807-836.
Health Canada (2024). Purpose information – Colour additive. In: Natural Health Products Ingredients Database – Drug and Health Products. Available at: https://webprod.hc-sc.gc.ca/nhpid-bdipsn/pfReq?id=3 [date modified: 2024-06-06].
Hoyer P, Flaws JA (2019). Toxic responses of the endocrine system (ch 20). In: Klaassen CD, ed. Casarett and Doull's Toxicology: The Basic Science of Poisons, 9th ed. New York: McGraw-Hill, pp. 977-1002.
Krishan M, Navarro L, Beck B, et al. (2021). A regulatory relic: After 60 years of research on cancer risk, the Delaney Clause continues to keep us in the past. Toxicol Appl Pharmacol 433:115779.
Lewandowski TA, Seeley MR, Beck BD (2004). Interspecies differences in susceptibility to perturbation of thyroid homeostasis: a case study with perchlorate. Regul Toxicol Pharmacol 39:348-362.
McClain RM (1989). The significance of hepatic microsomal enzyme induction and altered thyroid function in rats: Implications for thyroid gland neoplasia. Toxicol Pathol 17:294-306.
Ohshima M, Ward JM (1984). Promotion of N-methyl-N-nitrosourea-induced thyroid tumors by iodine deficiency in F344/NCr rats. J Natl Cancer Inst 73:289-296.
Ohshima M, Ward JM (1986). Dietary iodine deficiency as a tumor promoter and carcinogen in male F344/NCr rats. Cancer Res 46:877-883.
