The researchers observed a positive association between more frequent antibiotic use and CRC using the full dataset; the results were attenuated toward null when excluding antibiotics prescribed within two years of diagnosis. Excluding the two-year washout, the positive association was confined to the proximal colon in site-specific analyses (adjusted odds ratio for very high use versus no use, 1.17). An inverse association was seen for rectal cancer, which seemed to be driven by women. There was a positive association observed for quinolones and sulfonamides and/or trimethoprim with proximal colon cancer, while across antibiotics classes, a more general inverse association was seen for rectal cancer.
"There is absolutely no cause for alarm simply because you have taken antibiotics," a coauthor said in a statement. "The increase in risk is moderate and the [effect] on the absolute risk to the individual is fairly small."