Lead: Two new French cohort studies published on 7 January 2026 report associations between several widely used food preservatives and higher risks of multiple cancers and type 2 diabetes. Researchers working with the NutriNet-Santé cohort analyzed repeated, brand-specific 24-hour dietary records from more than 170,000 participants enrolled from 2009 and linked those reports to national health records. The cancer analysis followed about 105,000 cancer-free participants for up to 14 years; the diabetes analysis examined nearly 109,000 people without diabetes at baseline. The investigators found that specific preservatives — including sodium nitrite, potassium nitrate, sorbates, metabisulfites and several acetate forms — were associated with statistically higher risks for certain cancers and for developing type 2 diabetes.
Key takeaways
- The BMJ cancer study followed ~105,000 participants (free of cancer in 2009) for up to 14 years using repeated 24-hour, brand-specific dietary records.
- Researchers examined 58 preservatives overall, focused on 17 consumed by ≥10% of participants; six preservatives showed positive cancer associations despite US FDA GRAS status.
- Sodium nitrite was linked with a 32% higher risk of prostate cancer; potassium nitrate with a 22% higher risk of breast cancer and 13% higher risk of all cancers.
- Sorbates (notably potassium sorbate) correlated with a 26% higher breast cancer risk and 14% higher risk of all cancers; potassium metabisulfite showed ~20% higher breast cancer risk and 11% higher overall cancer risk.
- Acetates and acetic acid were associated with 25% and 12% higher risks respectively for certain cancers; sodium erythorbate/erythorbates were linked to elevated breast and overall cancer incidence.
- The Nature Communications diabetes analysis of ~109,000 participants found 12 of 17 examined preservatives associated with nearly a 50% higher risk of developing type 2 diabetes for highest consumers.
- Five preservatives (potassium sorbate, potassium metabisulfite, sodium nitrite, acetic acid and sodium acetate) overlapped as risks for both cancer and type 2 diabetes; calcium propionate was additionally linked to diabetes risk.
- Both studies adjusted for physical activity, smoking, alcohol, medication and multiple dietary confounders; authors stress observational design and recommend replication.
Background
The NutriNet-Santé cohort, started in 2009, collects web-based, brand-specific 24-hour dietary recalls and links them with French national health records to track long-term outcomes. That design allows investigators to estimate additive-specific intake rather than relying on generic categories of processed foods, by mapping brand names to ingredient lists and assigned additive content. Food preservatives are widely used across Europe and the United States to reduce microbial spoilage, extend shelf life and preserve color or texture; many are classified by regulators as GRAS (generally recognized as safe) or hold specific maximum use levels.
Regulatory assessments historically combine toxicology, animal studies and limited human exposure data; they rarely have large prospective cohorts with repeated, brand-level consumption data. Meanwhile, public-health bodies such as the World Health Organization have classified processed meat as a carcinogen on the basis of nitrate/nitrite chemistry and cooking-related nitrosation products, creating a precedent for concerns about additive-linked risks. At the same time, preservatives can come from natural fermentation (acetates) or be synthesized (nitrites, sorbates), and their biological effects may vary by chemical form, dose and the food matrix in which they are consumed.
Main event
The cancer paper, published in The BMJ on 7 January 2026, began with a pool of about 105,000 adults who completed frequent 24-hour brand-specific questionnaires and were cancer-free at baseline. Investigators quantified intake of 58 preservatives, narrowed their primary analysis to 17 additives consumed by at least 10% of participants, and used multivariable models to estimate associations with site-specific and overall cancer incidence through linkage with national health databases over up to 14 years.
The authors report six preservatives associated with elevated cancer risk: sodium nitrite, potassium nitrate, sorbates (particularly potassium sorbate), potassium metabisulfite, acetates and acetic acid. They also found that sodium erythorbate and related erythorbates were linked to higher breast and overall cancer incidence. Effect sizes varied by preservative and cancer site; for example, sodium nitrite was associated with a 32% greater prostate cancer risk in the highest consumers compared with the lowest.
Separately, the Nature Communications study analyzed nearly 109,000 NutriNet-Santé participants without type 2 diabetes at baseline and examined the same panel of preservatives for incident diabetes. Twelve of 17 preservatives showed positive associations with type 2 diabetes — with the highest consumers facing an approximately 49% higher probability of developing diabetes for five overlapping preservatives, and other additives (including some antioxidant forms) linked to roughly 42% higher risk.
Both research teams report extensive confounder adjustment (lifestyle factors, medication use, other dietary additives and preservatives from natural sources) and sensitivity analyses. The papers emphasize statistical associations rather than proven causation and call for experimental work, mechanistic studies and external cohort replication to corroborate the findings.
Analysis & implications
Biological plausibility for these associations is supported by experimental literature referenced by the authors: some preservatives can alter gut microbiota, generate oxidative stress, or modulate inflammatory pathways in animal and cell models. For nitrites and nitrates, chemistry that leads to nitrosamine formation has long been a toxicological concern, particularly in cured meats and high-temperature cooking. For sulfites and sorbates, possible effects on microbial metabolism and mucosal inflammation are hypothesized but not definitively proven in humans.
From a regulatory perspective, these results raise questions about safety assessments that treat additives in isolation or assume that concentrations permitted under GRAS/authorized limits are risk-free in long-term, cumulative exposure scenarios. The studies do not overturn enshrined regulatory decisions but suggest that population-level, long-duration exposure merits re-evaluation, especially for additives present across many packaged products.
For food manufacturers, the research could accelerate reformulation efforts already underway in some markets, shifting toward fewer synthetic additives, alternate preservation methods, or clearer labeling about additive content and function. For clinicians and public-health communicators, the consistent public-health message remains to prioritize minimally processed, plant-forward diets; these findings add a potential additive-specific rationale to longstanding dietary guidance.
Economically, tighter regulation or reformulation could increase costs for producers and potentially prices for consumers, but the public-health trade-off—if causal effects are later established—could justify preventive regulation. Policymakers will need to balance industry transition costs with precautionary measures and fund further research into dose thresholds, vulnerable populations and interactive effects among additives.
Comparison & data
| Preservative | Common uses | Reported cancer risk (highest vs lowest) | Reported diabetes risk (highest vs lowest) |
|---|---|---|---|
| Sodium nitrite | Processed meats (bacon, ham, deli) | +32% prostate cancer | Included among five preservatives → ~+49% diabetes |
| Potassium nitrate | Cured products, some processed foods | +22% breast; +13% all cancers | Linked (overlap with diabetes signals) |
| Potassium sorbate | Wine, baked goods, cheeses, sauces | +26% breast; +14% all cancers | Included in ~+49% diabetes group |
| Potassium metabisulfite | Winemaking, brewing | +20% breast; +11% all cancers | Included in ~+49% diabetes group |
| Acetates / acetic acid | Fermented foods, vinegar, sauces | Acetates +25% breast; acetic acid +12% all cancers | Included in ~+49% diabetes group |
The table summarizes the main preservatives highlighted in both papers. These figures represent relative risk increases reported for highest versus lowest reported consumption categories and do not convey absolute risk or baseline incidence. Absolute risk differences depend on age, sex, baseline disease rates and cumulative exposure and were not uniform across subgroups.
Reactions & quotes
Leading investigators and external experts urged caution while noting the public-health importance of the findings.
These are important initial results because the additives assessed are widespread in European, American and global food supplies; however, the associations require confirmation in other cohorts and mechanistic studies before policy shifts.
Mathilde Touvier — NutriNet-Santé principal investigator (paraphrased)
Context: Touvier highlighted the study’s novel use of brand-level dietary data and the need for replication; she also noted that the team adjusted for many potential confounders but cannot infer causality from observational data alone.
The concern raised about preservatives adds to reasons to favor whole, minimally processed plant foods for personal and public health.
Dr. David Katz — preventive medicine specialist, external commentator (paraphrased)
Context: Katz, not involved in the research, emphasized practical dietary advice while acknowledging the studies’ potential regulatory implications.
These data support reassessing regulations governing additive use to better protect consumers, especially given long-term, population-wide exposure.
Anaïs Hasenböhler — first author, Nutritional Epidemiology Research Team (paraphrased)
Context: Hasenböhler noted the papers are the first large prospective analyses linking specific preservatives to cancer and diabetes and called for regulatory review.
Unconfirmed
- These studies establish statistical associations but do not prove causation; alternative unmeasured factors could partly explain the links.
- Exact dose thresholds and safe intake levels for the implicated preservatives are not determined by these analyses.
- Potential interactions among multiple additives and between additives and food matrices remain incompletely characterized.
Bottom line
Two large prospective analyses from the NutriNet-Santé cohort identify consistent associations between several commonly used food preservatives and elevated risks of selected cancers and type 2 diabetes. The results are notable because some implicated preservatives are widely permitted under current regulatory frameworks, but they do not by themselves establish causality. Policymakers, regulators and manufacturers should consider these findings as a signal warranting further confirmation, mechanistic research and, where appropriate, precautionary reappraisal of additive approvals and labeling standards.
For consumers, the immediate, practical takeaway aligns with established dietary guidance: favor fresh, minimally processed foods and diverse plant-based choices to reduce cumulative exposure to multiple industrial additives. Researchers should prioritize replication in independent cohorts, randomized or mechanistic studies to test causality, and work to quantify absolute risk differences across demographic groups.
Sources
- CNN (news report summarizing the studies) — (news media)
- The BMJ (cancer study published in journal) — (peer-reviewed journal)
- Nature Communications (diabetes study published in journal) — (peer-reviewed journal)
- NutriNet-Santé cohort (study cohort information) — (academic/official cohort resource)
- World Health Organization (processed meat and cancer background) — (international public-health agency)