Author’s note: Two physician friends came to dinner this week. Michael, a pediatrician, asked about the disturbing rise in colon cancer among young people. I didn’t want to monopolize a dinner party with a soapbox launch, nor evade the question, so I gave him the shortest answer I could: “Pesticides.”

But a one-word answer to a disease increasingly affecting young adults is neither sufficient nor particularly good dinner conversation. So, Michael, this is the answer I wanted to give you.

What follows is not a declaration that one chemical, one food, or one exposure “causes” early onset colorectal cancer. The mechanisms remain under active investigation. Rather, this is an attempt to connect findings scattered across several scientific disciplines: environmental toxicology, microbiome research, cancer biology, immunology, agriculture, and pediatrics. When those pieces are placed beside one another, a troubling biological pathway begins to emerge, and it may begin much earlier than we thought.

The Patient I Haven’t Forgotten

I remember the first time I saw it. She was 27 years old, a runner and a graduate student in environmental science, ironically enough. She came to see me after six months of abdominal pain, intermittent rectal bleeding, and unexplained weight loss that had been attributed to “probably IBS” and “maybe stress.” By the time someone ordered a colonoscopy, she had a fungating mass in her sigmoid colon and liver metastases on CT.

She asked me why this was happening to her. She had no family history. She ate well, although not perfectly. She exercised. She was thin. She didn’t smoke. She had seemingly done everything right. I didn’t have an answer for her then, and I still don’t have a complete answer. No one does. But today we have clues that weren’t available to us then, and some of the most important points are not to adulthood, but to childhood.

The Data We Can No Longer Ignore

Early-onset colorectal cancer, generally defined as colorectal cancer diagnosed before age 50, has been rising for decades. The phenomenon is particularly disturbing because many cases occur in people without an obvious hereditary cancer syndrome. Something in our environment has changed. Genes do not change across a population in 30 years. Environments do.

That does not mean every environmental change is causal. It does mean we should be looking seriously at what changed in the food, chemical, microbial, and metabolic environments of the generations now developing these cancers. Increasingly, researchers are doing exactly that.

At the 2026 American Association for Cancer Research annual meeting, investigators reported an analysis of 108,315 deaths from early-onset colorectal cancer in the United States between 1989 and 2023. They compared mortality with county-level agricultural glyphosate use and found a statistically significant dose-response association: greater estimated glyphosate use was associated with progressively higher early-onset colorectal cancer mortality. The highest-exposure counties had approximately a 10% greater adjusted mortality rate than the lowest-exposure counties.

This study does not prove that glyphosate caused those cancers. County-level pesticide application is not the same thing as measuring an individual’s lifetime exposure, and the investigators themselves called for studies incorporating incidence, residential histories, and life-course exposure. But a dose-response signal involving the most heavily used herbicide in American agriculture deserves considerably more than a shrug.

Then came another clue. A 2026 study in Nature Medicine used DNA methylation signatures as biological proxies for environmental exposures in people with early versus late-onset colorectal cancer. Among numerous environmental factors examined, the strongest and most consistently replicated pesticide signal involved picloram. Population-level analyses subsequently found an association between county picloram use and early onset colorectal cancer incidence. Signals involving several other pesticides, including glyphosate and atrazine, were also reported.

Again, association is not causation. But when independent research begins pointing toward environmental chemical exposures in a cancer whose incidence is inexplicably increasing in younger generations, dismissal is no longer an adequate scientific response. We should be asking why.

The Gut Microbiome: Where the Story May Begin

To understand why pesticides might matter to colorectal cancer, we have to look at what actually lives in the colon. The human gastrointestinal tract contains an enormous ecosystem of bacteria, fungi, viruses, and other organisms collectively known as the microbiome. These organisms participate in digestion, vitamin production, epithelial integrity, immune education, metabolism, and inflammatory regulation.

Among the important members of the early life microbiome are Bifidobacterium species. These bacteria ferment dietary substrates and produce metabolites including acetate and other short-chain fatty acids. Those microbial metabolites help maintain intestinal barrier function and influence communication among the microbiome, intestinal epithelium, immune system, and nervous system.

The intestine, in other words, is not merely a digestive tube. It is an immune organ, a neurological organ, and one of the largest interfaces between the human body and the external environment.

One important regulatory system operating at this interface is the cholinergic anti-inflammatory pathway. Acetylcholine signaling through the alpha-7 nicotinic acetylcholine receptor, or α7nAChR, can restrain inflammatory cytokine production and help regulate excessive immune activation. It functions as one component of the body’s inflammatory braking system.

The question is not whether Bifidobacterium is the sole source of acetylcholine; it isn’t. Nor does the disappearance of one bacterial genus simply turn off the α7 receptor. The more scientifically interesting question is what happens when chronic environmental exposures alter the microbial metabolites, epithelial signals, immune pathways, and neural communication that collectively regulate inflammation in the developing gut.

Glyphosate belongs in that discussion. It targets the shikimate pathway, which human cells lack but many microorganisms possess. Laboratory studies have demonstrated that bacterial species differ in their susceptibility to glyphosate and glyphosate-based formulations. Precisely what chronic dietary exposure does to the human microbiome, particularly during infancy and childhood, remains inadequately characterized. That knowledge gap should concern us, not reassure us, because another microbial discovery has dramatically changed the early onset colorectal cancer conversation.

Colibactin: The Childhood Clue

In 2025, researchers published a remarkable study in Nature examining colorectal cancer genomes from 981 patients across eleven countries. They were looking for mutational signatures, the molecular fingerprints left behind by processes that damage DNA. One fingerprint stood out: colibactin.

Colibactin is a genotoxin produced by certain bacteria carrying specific genetic machinery, including some strains of E. coli. It can damage DNA in colon epithelial cells and leave behind characteristic mutational signatures known as SBS88 and ID18. Those signatures were dramatically enriched in younger colorectal cancer patients. Colibactin-associated signatures were 3.3 times more common in colorectal cancers diagnosed before age forty than in cancers diagnosed after age seventy.

But the finding that stopped me as a pediatrician was not simply the magnitude of the association. It was the timing. Prior work suggests that colibactin mutagenesis can occur during the first decade of life, and the investigators proposed that an early burst of DNA damage could give affected colon cells what amounts to a decades-long head start toward malignancy. In some tumors, colibactin-associated mutations affected APC, one of the critical tumor-suppressor genes involved early in colorectal carcinogenesis.

Consider the implication: the seeds of colon cancer diagnosed at 35 may have been planted in childhood. That possibility changes the entire conversation about prevention.

What Determines Which Bacteria Win?

The next question is obvious: why would colibactin-producing organisms gain a foothold in one child’s intestine and not another’s? We don’t yet know the complete answer, but microbial ecology matters. Birth mode matters. Breastfeeding matters. Antibiotic exposure matters. Diet matters. Ultra-processed food matters. Fiber matters. The intestinal environment matters. Environmental chemicals capable of altering microbial communities deserve investigation as well.

This is where I believe the glyphosate question becomes particularly important. I am not arguing that science has demonstrated a simple linear sequence in which glyphosate produces colibactin and colibactin produces colon cancer. It hasn’t. I am arguing that we now have enough pieces to ask whether chronic exposure to antimicrobial agricultural chemicals can alter the developing intestinal ecosystem in ways that reduce microbial resilience and create ecological opportunities for genotoxic organisms.

That is a testable hypothesis, and considering what we now know about colibactin and early-life DNA damage, it is one we should be testing urgently.

Why Call This the GMO-to-Cancer Pathway?

An important distinction needs to be made here: glyphosate is not a GMO, and the argument is not that a genetically engineered gene directly causes colorectal cancer. The connection is agricultural and ecological.

Glyphosate use increased dramatically following the introduction of glyphosate-resistant genetically engineered crops beginning in the mid-1990s. That technology made it possible to spray glyphosate directly over vast acreage of commodity crops engineered to survive the herbicide. Corn and soy produced within this system became foundational ingredients in the ultra-processed Western diet. Glyphosate is also used on some non-GMO crops, including in certain pre-harvest applications.

When I use the phrase “GMO-to-cancer pathway,” therefore, I am describing a larger agricultural system in which genetically engineered herbicide-tolerant crops facilitated intensive herbicide use and increased opportunities for human environmental and dietary exposure. The hypothesis I am asking us to investigate is whether those exposures can contribute to microbiome disruption, chronic inflammation and microbial genotoxicity, ultimately increasing susceptibility to carcinogenesis.

Not every link in that pathway has been proven. But enough of the individual links now have evidence behind them that the pathway deserves serious investigation.

A Second Question: Spike Biology and the Tumor Microenvironment

There is another emerging area that warrants investigation, but it needs to be kept scientifically distinct from the pesticide evidence. The SARS-CoV-2 spike protein interacts with numerous cellular pathways, and experimental studies have examined potential interactions between spike and nicotinic acetylcholine receptors, including α7nAChR, a receptor involved in cholinergic regulation of inflammation. Other laboratory studies have reported effects of SARS-CoV-2 proteins on signaling pathways relevant to cancer biology.

A 2023 study, for example, found that SARS-CoV-2 spike and nucleocapsid peptides could induce TGF-β1 signaling and epithelial-mesenchymal transition in colon cancer cells in vitro. EMT matters because it is one of the biological programs through which epithelial cancer cells can acquire more migratory and invasive characteristics.

This is intriguing biology, but it is not evidence that COVID infection or COVID vaccination causes colorectal cancer. Nor does an experiment performed in cultured cancer cells tell us what happens in a human being over decades. What it does raise is a legitimate research question: Could spike-related signaling modify inflammation or tumor behavior in individuals who already harbor precancerous or malignant cells?

There is another distinction that must be maintained. SARS-CoV-2 infection and mRNA vaccination are not biologically identical exposures simply because both can result in exposure to spike protein. Dose, location, duration, immune context, and accompanying viral biology differ. The appropriate scientific response is therefore neither “case closed” nor “nothing to see here.” It is investigation.

If we are witnessing an unexplained increase in aggressive cancers among younger adults, every biologically credible environmental and immunological hypothesis should be available for study without political or institutional interference in either direction.

The Larger Problem: Our Scientific Silos

This brings me to what may be the real problem. The clues are scattered across disciplines that rarely speak to one another. Cancer researchers study mutations while microbiome researchers study dysbiosis. Toxicologists study pesticides while immunologists study inflammatory signaling. Agricultural scientists study crop systems while pediatricians see the earliest manifestations of disturbed gastrointestinal and immune function.

We need these disciplines sitting at the same table. If early-onset colorectal cancer is partly the downstream consequence of environmental exposures beginning in childhood, prevention cannot begin when a thirty-five-year-old walks into an oncologist’s office. It must begin decades earlier.

Forests, Farms, and the Exposure We Cannot Choose

There is also an important distinction between voluntary and involuntary exposure. A family can choose organic oatmeal, filter its drinking water, or reduce ultra-processed foods. Communities cannot individually opt out when herbicides are applied across agricultural landscapes, utility corridors, timberlands, or public forests connected to their watersheds.

That is why the emerging picloram research deserves particular attention. The 2026 Nature Medicine study found the strongest pesticide association with picloram, an herbicide used for control of broadleaf plants and woody vegetation. The authors concluded that their findings highlight the potential importance of the exposome in early onset colorectal cancer.

This does not prove that forestry spraying causes colon cancer. It does mean that environmental monitoring, biomonitoring, watershed protection, and transparent reporting of pesticide applications are legitimate public-health issues. Children do not choose their exposures, and if some of the mutations initiating adult cancer are acquired during childhood, that fact carries profound implications for environmental policy.

What Must Change

We need serious research into pesticide exposure and early onset colorectal cancer, and we need to move beyond another twenty years of ecological correlations. Prospective research should incorporate individual exposure measurements, residential and occupational histories, pesticide metabolites, microbiome sequencing, dietary patterns, tumor genomics, and longitudinal health outcomes. If the signal is real, modern molecular tools give us the ability to find it.

We also need to study the developing microbiome as though it matters. Children are not small adults. Their microbial ecosystems, immune systems, intestinal barriers, metabolic pathways, and nervous systems are developing simultaneously. An exposure that produces little measurable effect in a forty-year-old may have very different consequences during infancy or childhood.

At the same time, we should dramatically expand organic and regenerative agriculture. If we can produce food while reducing dependence on herbicides and pesticides, improving soil health, protecting watersheds, and decreasing unnecessary chemical exposure, why wouldn’t we? Public policy should help farmers make that transition rather than financially locking them into chemical-dependent monocultures.

Landscape-scale pesticide applications also deserve far greater scrutiny. The burden should not fall on communities to prove decades later that an exposure harmed them. When chemicals are dispersed into shared environments, precaution, monitoring, transparency, and meaningful community involvement should be the minimum standard.

Finally, medicine needs intellectual courage. We should be able to investigate pesticides without being labeled anti-agriculture, examine pharmaceutical products without being labeled anti-medicine, and acknowledge when an intriguing mechanism remains a hypothesis rather than pretending we know more than we do. Science advances by asking uncomfortable questions and then designing experiments capable of proving us wrong.

The Pediatrician’s View

I did not set out to write about colon cancer. I am a pediatrician, and for decades I have watched children arrive with gastrointestinal dysfunction, immune dysregulation, neurodevelopmental challenges, metabolic disease, and inflammatory conditions that were once far less common in pediatric practice. I have become increasingly convinced that we make a mistake when we examine each diagnosis in isolation.

The intestine sits at the intersection of the external environment and the internal terrain. Food and chemical residues arrive there. Microbes live there. Immune cells patrol there. The nervous system communicates there. The intestinal epithelium continually regenerates there. And now we have evidence that a bacterial genotoxin may leave DNA damage in the colon during childhood that remains visible in a cancer appearing decades later.

That should change how we think about cancer prevention. Perhaps the most important question is not simply why a thirty-five-year-old is developing colon cancer. Perhaps we also need to ask what happened in that person’s gut when they were five.

That is the question I want pediatricians, oncologists, microbiologists, toxicologists, farmers, regulators, and parents asking together. We do not yet have the entire causal chain, but we have enough pieces that refusing to investigate how they may fit together is no longer scientifically satisfying.

Our children deserve that investigation. So do the young adults walking into oncology offices with diseases appearing decades earlier than expected, and the communities living amid agricultural and environmental chemical exposures they did not choose.

Michael, next time we have dinner, perhaps we can talk about something lighter. But this is the answer I wanted to give you. It is also the answer I wish I could have given that twenty-seven-year-old patient when she looked at me and asked, “Why?”

References

1. Lin J, Cortez M, Nondin C, et al. Agricultural glyphosate use and early-onset colorectal cancer mortality in the United States. Cancer Research. 2026;86(7 Suppl):Abstract 6244. The analysis included 108,315 EOCRC deaths and reported a dose-response association between county-level glyphosate use and mortality.
2. Maas SCE, Baraibar I, Lemler L, et al. Epigenetic fingerprints link early-onset colon and rectal cancer to pesticide exposure. Nature Medicine. 2026;32:1827– doi:10.1038/s41591-026-04342-5. Picloram produced the strongest replicated association, while glyphosate also showed relatively robust county-level associations.
3. Díaz-Gay M, dos Santos W, et al. Geographic and age variations in mutational processes in colorectal cancer. Nature. 2025. doi:10.1038/s41586-025-09025-8. This is the key childhood/colibactin paper: 981 CRC genomes from 11 countries, enrichment of SBS88 and ID18 in EOCRC, 3.3-fold enrichment under age 40 versus over 70, and evidence that colibactin-related damage occurs early in tumor development.
4. Spaander MCW, Zauber AG, Syngal S, et al. Young-onset colorectal cancer. Nature Reviews Disease Primers. 2023;9:21. doi:10.1038/s41572-023-00432-7. Excellent general reference for the epidemiology and the importance of environmental and childhood exposures.
5. Hofseth LJ, Hebert JR, Chanda A, et al. Early-onset colorectal cancer: initial clues and current views. Nature Reviews Gastroenterology & Hepatology. 2020;17:352–364. doi:10.1038/s41575-019-0253-4. Particularly useful because the authors explicitly place the gut microbiota at the crossroads of multiple suspected EOCRC risk factors and discuss exposures occurring from conception through childhood.
6. Atchade AM, Williams JL, Mermelstein L, Nemesure B. Unraveling the complexities of early-onset colorectal cancer: a perspective on dietary and microbial influences. Frontiers in Public Health. 2024;12:1370108. doi:10.3389/fpubh.2024.1370108. Supports the argument for moving beyond obesity toward diet, antibiotics, processed foods and microbial ecology.
7. Ignácio AC, Guerra AMDR, de Souza-Silva TG, do Carmo MAV, Paula HAA. Effects of glyphosate exposure on intestinal microbiota, metabolism and microstructure: a systematic review. Food & Function. 2024;15(15):7757–7781. doi:10.1039/D4FO00660G. The review reports dysbiosis, permeability and intestinal structural changes across experimental studies, while also making clear the limitations of the evidence base.
8. Mathew SA, Muola A, Saikkonen K, Saloniemi I, Helander M, Puigbò Quantification of the potential impact of glyphosate-based products on microbiomes. Journal of Visualized Experiments. 2022;(179). doi:10.3791/63109.
9. Puigbò P, et al. Does glyphosate affect the human microbiota? Life. 2022. The genomic analysis found substantial variation in predicted glyphosate susceptibility across members of the human microbiome and emphasizes the need for empirical human studies.
10. Thakur BK, Malaise Y, Choudhury SR, et al. Dietary fibre counters the oncogenic potential of colibactin-producing Escherichia coli in colorectal cancer. Nature Microbiology. 2025;10:855– doi:10.1038/s41564-025-01938-4. It ties diet + microbiome + colibactin + carcinogenesis together and gives the article a constructive nutritional dimension.
11. Jans M, Kolata M, Blancke G, et al. Colibactin-driven colon cancer requires adhesin-mediated epithelial binding. Nature. 2024. doi:10.1038/s41586-024-08135-z. Strong mechanistic support that colibactin-producing coli can participate directly in colorectal carcinogenesis rather than simply being an associated microbial marker.
12. O’Brien BCV, Weber L, Hueffer K, Weltzin MM. SARS-CoV-2 spike ectodomain targets α7 nicotinic acetylcholine receptors. Journal of Biological Chemistry. 2023;299(5):104707. doi:10.1016/j.jbc.2023.104707.
13. Changeux JP, et al. A functional interaction between the Y674-R685 region of the SARS-CoV-2 spike protein and the human α7 nicotinic receptor. Molecular Neurobiology. 2022;59:6076–6090.
14. SARS-CoV-2 S and N protein peptides drive invasion abilities of colon cancer cells through TGF-β1 regulation. BBA Molecular Cell Research. 2023;1870:119541. doi:10.1016/j.bbamcr.2023.119541. This directly supports the in-vitro spike/TGF-β1/EMT discussion.
15. Cáceres-Chacón M, Martínez-Guzmán O, Haddock-Martínez HA, et al. Exposure to the herbicide glyphosate leads to inappropriate threat responses and alters gut microbial composition. Frontiers in Toxicology. 2025;7:1704231. doi:10.3389/ftox.2025.1704231. This is a newer experimental glyphosate-microbiome reference.