Introduction

For centuries, food processing had a straightforward purpose: to make foods safer, easier to digest and more stable (through cooking, fermentation, drying and preservation). This kind of processing alters food, but generally preserves an identifiable food matrix — fibres, structure, texture, and digestion kinetics.

Modern industrial eating introduced a clear break: ultra-processed foods (UPFs) are not merely processed — they are reassembled. They combine fractionated ingredients (isolates, modified starches, syrups), technological additives (emulsifiers, sweeteners, flavours), and intensive processes (extrusion, refining, recombination). The goal is not only nutritional: it is also technological (texture, shelf stability), economic (cost), and sensory (palatability).

Biologically, this changes the nature of the signals the body receives: speed of eating, energy density, impact on the microbiota, glycaemic load, repeated exposure to certain additives, newly formed processing contaminants, and contact materials. So it is not just a matter of calories — it is a matter of biological information perceived by the cell.

I — What do we mean by “ultra-processed foods” (and why the nuance matters)?

1) The NOVA classification: useful, but it must be used rigorously

UPFs correspond to Group 4 of the NOVA classification. NOVA is operationally useful: it helps identify a category of highly industrialised foods, often formulated from fractionated components and additives. But it does not replace a mechanistic analysis: not all foods within a given group are equivalent, and the biology runs through several pathways at once (matrix, density, additives, contaminants, eating behaviour) [6].

2) The key point: the “food matrix”

Two foods can display a similar nutritional profile (sugar/fat/salt), yet trigger different biological responses depending on structure: digestion speed, glycaemic response, satiety, colonic fermentation, exposure to newly formed compounds. This is one reason why reducing the debate to “macros” or “calories” often leads to clinical dead ends [5].

II — What the French scientific literature shows: robust associations in NutriNet-Santé

France has a particularly strong body of evidence thanks to NutriNet-Santé, a large prospective cohort led by teams from Inserm/INRAE/Cnam/Université Paris Cité and partners. Observational studies do not prove causality on their own, but they provide a coherent, reproducible and biologically plausible pattern of associations across multiple outcomes [1–4].

1) Cancer

Higher UPF consumption is associated with an increased risk of cancer in NutriNet-Santé [1]. The authors explicitly highlight that several dimensions of ultra-processing may be involved: nutritional composition, additives, contact materials, and newly formed contaminants.

2) Cardiovascular disease

In NutriNet-Santé, higher UPF intake is associated with a greater risk of cardiovascular disease (coronary and cerebrovascular) [2]. This strengthens the idea that ultra-processing acts beyond body weight alone, through inflammation, dyslipidaemia, metabolic dysregulation and vascular mechanisms.

3) Type 2 diabetes

In JAMA Internal Medicine, NutriNet-Santé reports an association between a higher dietary proportion of UPFs and increased risk of type 2 diabetes [3]. Here again, the signal reflects a dietary pattern and a metabolic environment — not a single isolated nutrient.

4) Mortality

Higher UPF consumption is associated with increased all-cause mortality risk in NutriNet-Santé [4]. This type of endpoint is particularly informative because it indirectly integrates multiple biological pathways.

III — The essential experimental link: “calories matched”, ultra-processed still drives overeating

One common critique of epidemiology is residual confounding: “people who eat more UPFs also have other risk factors.” That is why a controlled experimental trial matters.

In the NIH inpatient randomised trial, two diets — ultra-processed vs minimally processed — were designed to be “matched” on several presented nutritional parameters (calories offered, macronutrients, sugar, sodium, fibre). Result: the ultra-processed diet led to a spontaneous increase in intake (≈ +500 kcal/day), weight gain, and changes in satiety hormones within weeks [5]. In other words, ultra-processing influences eating behaviour and satiety physiology — independently of the idea that it is “just willpower”.

IV — Biological mechanisms: why ultra-processing disrupts the cell

The mechanisms are multifactorial. The strength of the topic is precisely that it is not driven by a single “culprit”, but by an accumulation of pro-inflammatory and pro-metabolic signals.

1) Energy density + eating speed: the biology of overconsumption

UPFs are often easier to chew, faster to eat and more hyper-palatable. This combination promotes quicker intake, a disconnect between ingestion and satiety signalling, and repeated exposure to glycaemic/insulin peaks. The controlled trial reinforces the point: the effect is not only behavioural — it is structural [5].

2) A depleted matrix: fewer fibres, less protective fermentation

Loss of fermentable fibres and bioactive compounds (polyphenols, plant structures) reduces production of protective microbial metabolites and weakens intestinal balance. This can contribute to low-grade inflammation and metabolic dysregulation [6].

3) Additives: the issue is not acute toxicity, but chronic, cumulative, systemic effects

Several recent French studies (CRESS-EREN / NutriNet-Santé) explore associations between exposure to certain additives and metabolic, cardiovascular and cancer risks [7–9].

• Emulsifiers and diabetes: associations between exposure to certain emulsifiers and risk of type 2 diabetes, consistent with experimental data on the microbiota and inflammation [8].
• Emulsifiers and cardiovascular risk: associations observed with cardiovascular outcomes [7].
• Emulsifiers and cancer: cohort data and related syntheses, consistent with biological plausibility via the microbiota, metabolome and inflammation [9].

4) Preservatives: a new French signal on cancer and type 2 diabetes

Recent NutriNet-Santé publications analysing exposure to several food preservatives report associations with incidence of certain cancers and type 2 diabetes [10,11], echoed by Inserm [12]. These findings require mechanistic confirmation, but they strengthen the overall coherence of a chronic-exposure model.

5) Newly formed contaminants and processing: what ANSES emphasises

ANSES highlights the complexity of the pathways involved: certain industrial processes (high temperatures, low moisture, Maillard reactions) can favour the formation of newly formed substances. UPFs need to be analysed at the level of processes and cumulative exposures — not solely through a nutrition label [6].

V — Why it’s not “just about weight”: the inflammatory axis and biological ageing

Even without dramatic weight gain, chronic exposure to UPFs may promote:

• low-grade inflammation (meta-inflammation)
• microbiota disruption and increased gut permeability
• glycaemic and insulin dysregulation
• functional oxidative stress
• a progressive loss of metabolic resilience

These mechanisms converge towards inflammaging and biological ageing, in line with observed findings on cancer, cardiovascular disease, diabetes and mortality [1–4].

VI — Clinical translation: how to use these data without caricature

The aim is not to demonise industry or promise total control, but to adopt a realistic, biologically coherent approach:

• risk is dose-dependent
• consistency matters more than exceptions
• gradual substitution is the central lever
• overall coherence matters more than perfection

VII — Cellular Nutrition: rebuilding coherent dietary information

From a Cellular Nutrition perspective, the central question becomes:
what biological signals does your diet send to your cells every day?

Reducing ultra-processing helps to:

• lower aggressive metabolic load
• restore the microbiota through fibre and diversity
• reduce certain repeated exposures (additives, processes)
• rebuild functional micronutrient density
• support mitochondrial resilience and inflammatory stability

This reading is aligned with institutional positions (ANSES) and with French scientific output from NutriNet-Santé [1–12].

Conclusion — Ultra-processed foods: a biological issue, not only a nutritional one

UPFs do not only pose a caloric problem. They alter food structure, eating speed, the gut ecosystem and metabolic signalling — and they increase repeated exposure to technological and chemical combinations.

The strength of the French data lies in internal coherence, reproducibility, and convergence with plausible biological mechanisms and experimental findings.

The most robust response is neither fear nor denial, but a progressive strategy to reduce exposure — grounded in biology, coherence and long-term sustainability.

Bibliography

[1] Fiolet, T. et al. (2018) Consumption of ultra-processed foods and cancer risk: results from NutriNet-Santé. BMJ, 360, k322.
Available at: https://www.bmj.com/content/360/bmj.k322
PubMed: https://pubmed.ncbi.nlm.nih.gov/29444771/

[2] Srour, B. et al. (2019) Ultra-processed food intake and risk of cardiovascular disease: prospective cohort study (NutriNet-Santé). BMJ, 365, l1451.
Available at: https://www.bmj.com/content/365/bmj.l1451
PubMed: https://pubmed.ncbi.nlm.nih.gov/31142457/

[3] Srour, B. et al. (2019) Ultraprocessed Food Consumption and Risk of Type 2 Diabetes Among Participants of the NutriNet-Santé Prospective Cohort. JAMA Internal Medicine.
Available at: https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2757497
PubMed: https://pubmed.ncbi.nlm.nih.gov/31841598/
NutriNet-Santé notice (FR): https://etude-nutrinet-sante.fr/article/view/329-Associations-entre-la-consommation-d%E2%80%99aliments-ultra-transform%C3%A9s-et-le-risque-de-diab%C3%A8te-de-type-2-dans-la-cohorte-NutriNet-Sant%C3%A9

[4] Schnabel, L. et al. (2019) Association Between Ultraprocessed Food Consumption and Risk of Mortality Among Middle-aged Adults in France. JAMA Internal Medicine, 179(4), 490–498.
Available at: https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2723626
PubMed: https://pubmed.ncbi.nlm.nih.gov/30742202/
NutriNet-Santé notice (FR): https://etude-nutrinet-sante.fr/article/view/230-Consommation-d%E2%80%99aliments-ultra-transform%C3%A9s-et-risque-de-mortalit%C3%A9-%3A-les-r%C3%A9sultats-issus-de-l%E2%80%99%C3%A9tude-NutriNet-Sant%C3%A9

[5] Hall, K.D. et al. (2019) Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake. Cell Metabolism, 30(1), 67–77.e3.
Available at: https://www.cell.com/cell-metabolism/fulltext/S1550-4131(19)30248-7
PDF: https://www.cell.com/cell-metabolism/pdf/S1550-4131(19)30248-7.pdf
PubMed: https://pubmed.ncbi.nlm.nih.gov/31105044/

[6] ANSES (2024) Avis relatif à la caractérisation et évaluation des impacts sur la santé de la consommation d’aliments dits ultratransformés (saisine 2022-SA-0155).
ANSES page: https://www.anses.fr/fr/content/avis-relatif-la-caracterisation-et-evaluation-des-impacts-sur-la-sante-de-la-consommation
Report (PDF FR): https://www.anses.fr/fr/system/files/NUT2022-SA-0155.pdf
Report (PDF EN): https://www.anses.fr/sites/default/files/NUT2022-SA-0155-EN.pdf?download=1

[7] Sellem, L., Srour, B. et al. (2023) Food additive emulsifiers and risk of cardiovascular disease: prospective cohort study (NutriNet-Santé). BMJ, 382, e076058.
Available at: https://www.bmj.com/content/382/bmj-2023-076058

[8] Salame, C., Srour, B. et al. (2024) Food additive emulsifiers and the risk of type 2 diabetes: prospective cohort study (NutriNet-Santé). The Lancet Diabetes & Endocrinology.
Available at: https://www.thelancet.com/journals/landia/article/PIIS2213-8587(24)00086-X/fulltext

[9] Sellem, L., Srour, B. et al. (2024) Food additive emulsifiers and cancer risk: prospective cohort study (NutriNet-Santé). PLOS Medicine.
Available at: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004338
PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC10863884/

[10] Hasenböhler, A., Debras, C. et al. (2026) Intake of food additive preservatives and incidence of cancer: prospective cohort study (NutriNet-Santé). BMJ, 392, e084917.
Available at: https://www.bmj.com/content/392/bmj-2025-084917
PubMed: https://pubmed.ncbi.nlm.nih.gov/41500678/

[11] Hasenböhler, A., Debras, C. et al. (2025) Associations between preservative food additives and type 2 diabetes incidence: prospective cohort study (NutriNet-Santé). Nature Communications.
Available at: https://www.nature.com/articles/s41467-025-67360-w

[12] Inserm (2026) Deux nouvelles études suggèrent une association entre la consommation de conservateurs et un risque accru de cancer et de diabète de type 2 (communiqué de presse).
Available at: https://presse.inserm.fr/deux-nouvelles-etudes-suggerent-une-association-entre-la-consommation-de-conservateurs-et-un-risque-accru-de-cancer-et-de-diabete-de-type-2/71653/