Microbiota and colorectal cancer — controlling for confounders challenges associations

Katrina Ray
DOI: https://doi.org/10.1038/s41575-024-00947-8
2024-05-23
Nature Reviews Gastroenterology & Hepatology
Abstract:Controlling for confounders in quantitative microbiome data challenges the robustness of some reported associations between the gut microbiota and colorectal cancer (CRC) stages, according to a recent study. The faecal microbiota of 589 patients at different CRC stages was analysed and compared with up to 15 published studies (totalling 4,439 patients and controls). Quantitative microbiome profiling combined with rigorous confounder control identified transit time, faecal calprotectin levels (a marker of intestinal inflammation) and BMI as primary microbial covariates, which superseded variance explained by CRC diagnostic groups (healthy, adenoma and carcinoma). Notably, when controlling for these covariates, well-known CRC microbiome members (such as Fusobacterium nucleatum ) were not associated with CRC diagnostic groups. Instead, statistically significant associations were observed with a range of different microorganisms, such as Anaerococcus vaginalis and Dialister pneumosintes , among others. The findings highlight the important consideration of confounders in association studies.
gastroenterology & hepatology
What problem does this paper attempt to address?