The different processes that damage DNA have sequence preferences for the direct change in the DNA (e.g. G-->C versus G-->T), as well as the sequence around the damage. Smoking causes very very different signatures of DNA damage than, say, damage from UV from being in the sun.
So every cell in our body carries with it a (noisy) history of the mutagenic processes it has been exposed to.
This paper found some differences in these mutational signatures of later colon cancers, and attributed them to colibactin, a mutagen produced by bacteria.
It remains to be seen what percentage of the increase in under-50 colon cancers this would explain, but it's an additional risk factor that didn't make it onto the chart, likely because it's not coming from standard epidemiological analysis, and instead from the world of molecular analysis.
A recent finding from last year looked at the mutational signatures in colon cancer in those under 50:
https://www.nature.com/articles/s41586-025-09025-8
The different processes that damage DNA have sequence preferences for the direct change in the DNA (e.g. G-->C versus G-->T), as well as the sequence around the damage. Smoking causes very very different signatures of DNA damage than, say, damage from UV from being in the sun.
So every cell in our body carries with it a (noisy) history of the mutagenic processes it has been exposed to.
This paper found some differences in these mutational signatures of later colon cancers, and attributed them to colibactin, a mutagen produced by bacteria.
It remains to be seen what percentage of the increase in under-50 colon cancers this would explain, but it's an additional risk factor that didn't make it onto the chart, likely because it's not coming from standard epidemiological analysis, and instead from the world of molecular analysis.