ICU and gut microbiota

Background: Critical illness is associated with an altered gut microbiota (GM), known as dysbiosis, and this might be associated with adverse outcomes. However, previous research has yielded conflicting results, often focused on single time points without evaluating dynamic changes. GM research is challenging and lacks surrogate markers. Objectives: This study aimed to investigate early changes in GM in patients admitted to the ICU and assess its association with 60-day mortality. Furthermore, we aimed to determine whether fecal pH could serve as a simplified biomarker for GM alterations. Methods: This monocentric prospective study was conducted at Geneva University Hospital, Switzerland. Fecal samples were collected at two distinct time points: S1, representing the first stool passed upon ICU admission, and S2, the first stool passed at least 48 hours after the initial stool. GM analysis involved full-length 16S rRNA gene sequencing, with alpha diversity assessed using the Shannon index at the zOTU level. Changes in GM over time were analyzed using the Wilcoxon rank-signed test, while associations between GM diversity changes and mortality and fecal pH were evaluated using Bayesian joint models. Results: Ninety-six patients were included, and 24 deaths occurred within 60 days. GM diversity decreased significantly between S1 and S2 (p<0.01). The Shannon index decreased daily by -0.1 (CrI -0.16 to -0.04, p< 0.01). A decrease in the Shannon Index over time was associated with a 2.13-fold increased hazard of death (95% CrI, 1.06 to 4.71, p=0.032). However, no association was found between changes in fecal pH and changes in the Shannon Index over time. Conclusion: Our study found significant variation of GM among ICU patients upon admission, which decreased over time, and was associated with an increased 60-day mortality risk. Stool pH did not emerge as a suitable surrogate marker. Future interventions should target the early ICU days, a critical period for GM modification, and work on developing a reliable marker for clinical use.