Agreement: I Agree Body: Dear Editor The findings of Hasenböhler and colleagues [1] are significant not only for identifying specific preservatives but also for their broader implications for the structure of food additive regulation. Beyond established concerns regarding biological plausibility and acceptable daily intake thresholds, two structural issues, regrettable substitution and the mixture problem, are likely to determine whether these results translate into effective consumer protection. The history of chemical regulation demonstrates a recurring pattern. When bisphenol A was scrutinized, manufacturers substituted it with bisphenol S, F, and related analogues, structurally similar compounds with comparable endocrine-disrupting potential but limited epidemiological data, a phenomenon now formally termed "regrettable substitution" [2]. If sodium nitrite, potassium sorbate, or sodium erythorbate are restricted, industry is likely to respond by substituting chemically related but less-studied alternatives. Hasenböhler's observation that sodium erythorbate (E316) exhibits associations entirely absent from its stereoisomer, ascorbic acid (E300), highlights that minor structural differences can result in significant biological consequences that pre-market evaluation often fails to detect [1]. Re-evaluating individual E-numbers in isolation enables manufacturers to replace one inadequately characterized molecule with another. The problem of mixtures further complicates this issue. Almost all participants in NutriNet-Santé were simultaneously exposed to multiple preservatives (99.7% consumed at least one) [1]; yet, acceptable daily intake levels have been established for individual compounds that were tested in isolation. In the cytotoxicity studies cited by the authors, combinations were tested, and effects were identified that could not be predicted based on individual risks [3]. In a real diet, these compounds exist as a mixture alongside emulsifiers, sweeteners, and coloring agents, each of which can potentially influence microbiome composition, intestinal permeability, and inflammatory status [4]. An umbrella review conducted by Lane and his colleagues has linked exposure to ultra-processed foods to 32 adverse health outcomes [5], Nevertheless, Hassenboehler’s finding that associations with preservatives persist even after adjusting for the proportion of ultra-processed foods, suggests that these compounds harbor independent signals worthy of analysis, the scope of which extends far beyond the confines of the UPF label [1]. Therefore, to translate these findings into safety measures, merely re-evaluating individual additives is not sufficient. This requires a regulatory framework that assesses structural classes rather than individual molecules, mandates the testing of mixtures for toxicity prior to market entry, and establishes a system for post-market surveillance, akin to pharmacovigilance. Without such reforms, banning the preservatives currently under scrutiny could simply shift the risk toward similar substances of tomorrow that have yet to be studied, thereby perpetuating uncertainty and leaving the status quo of the industry unchanged. References 1. Hasenböhler, A., et al., Intake of food additive preservatives and incidence of cancer: results from the NutriNet-Santé prospective cohort. Bmj, 2026. 392: p. e084917. 2. Trasande, L., Exploring regrettable substitution: replacements for bisphenol A. Lancet Planet Health, 2017. 1(3): p. e88-e89. 3. Recoules, C., et al., Evaluation of the toxic effects of food additives, alone or in mixture, in four human cell models. Food Chem Toxicol, 2025. 196: p. 115198. 4. Whelan, K., et al., Ultra-processed foods and food additives in gut health and disease. Nat Rev Gastroenterol Hepatol, 2024. 21(6): p. 406-427. 5. Lane, M.M., et al., Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. Bmj, 2024. 384: p. e077310. No competing Interests: Yes The following competing Interests: Electronic Publication Date: Sunday, May 10, 2026 - 10:16 AI use: Yes I have used AI Highwire Comment Subject: Intake of food additive preservatives and incidence of cancer: results from the NutriNet-Santé prospective cohort AI use details: Date search and language correction Workflow State: Released Full Title: Regrettable substitution: structural gaps in food additive regulation Highwire Comment Response to: Intake of food additive preservatives and incidence of cancer: results from the NutriNet-Santé prospective cohort Check this box if you would like your letter to appear anonymously:: Last Name: Rahaman First name and middle initial: Abdul Email: rahaman_knabdul@ymail.com Address: Guangdong Key Laboratory of Food Intelligent Manufacturing Foshan University, Foshan, Guangdong, China Occupation: Assistant Professor Other Authors: Professor Xin-An Zeng, Dr. Ankita Kumari Affiliation: School of Food Science and Engineering, Foshan University, Nanhai, Foshan, 528225, Guangdong, China BMJ: Additional Article Info: Rapid response
Regrettable substitution: structural gaps in food additive regulation
