Safety, feasibility, and impact on the gut microbiome of kefir administration in critically ill adults

Abstract

Background: Dysbiosis of the gut microbiome is frequent in the intensive care unit (ICU), potentially leading to a heightened risk of nosocomial infections. Enhancing the gut microbiome has been proposed as a strategic approach to mitigate potential adverse outcomes. While prior research on select probiotic supplements has not successfully shown to improve gut microbial diversity, fermented foods offer a promising alternative. In this open-label phase I safety and feasibility study, we examined the safety and feasibility of kefir as an initial step towards utilizing fermented foods to mitigate gut dysbiosis in critically ill patients.

Methods: We administered kefir in escalating doses (60 mL, followed by 120 mL after 12 h, then 240 mL daily) to 54 critically ill patients with an intact gastrointestinal tract. To evaluate kefir's safety, we monitored for gastrointestinal symptoms. Feasibility was determined by whether patients received a minimum of 75% of their assigned kefir doses. To assess changes in the gut microbiome composition following kefir administration, we collected two stool samples from 13 patients: one within 72 h of admission to the ICU and another at least 72 h after the first stool sample.

Results: After administering kefir, none of the 54 critically ill patients exhibited signs of kefir-related bacteremia. No side effects like bloating, vomiting, or aspiration were noted, except for diarrhea in two patients concurrently on laxatives. Out of the 393 kefir doses prescribed for all participants, 359 (91%) were successfully administered. We were able to collect an initial stool sample from 29 (54%) patients and a follow-up sample from 13 (24%) patients. Analysis of the 26 paired samples revealed no increase in gut microbial α-diversity between the two timepoints. However, there was a significant improvement in the Gut Microbiome Wellness Index (GMWI) by the second timepoint (P = 0.034, one-sided Wilcoxon signed-rank test); this finding supports our hypothesis that kefir administration can improve gut health in critically ill patients. Additionally, the known microbial species in kefir were found to exhibit varying levels of engraftment in patients' guts.

Conclusions: Providing kefir to critically ill individuals is safe and feasible. Our findings warrant a larger evaluation of kefir's safety, tolerability, and impact on gut microbiome dysbiosis in patients admitted to the ICU.

Trial registration: NCT05416814; trial registered on June 13, 2022.

Keywords: Critical illness; GMWI; Gut microbiome; ICU; Kefir; Probiotics.

Fig. 1

Patient selection and stool sample collection for gut microbiome analysis. Of an initial 722 ICU admissions screened, 54 patients were enrolled in the current study based on our criteria. For gut microbiome analysis, stool samples were collected from patients after kefir administration; however, 25 of the 54 participants did not have bowel movements during their ICU stay and were unable to provide a stool sample. For the remaining 29 patients, an initial stool sample was collected within 72 h of ICU admission (from all 29 patients), with a subsequent sample gathered after kefir administration (from 13 patients), contingent on bowel activity

Fig. 2

Stacked bar plots showing the relative taxonomic abundances in the gut microbiomes of 29 ICU patients at two timepoints: T1 (within 72 h of ICU admission) and T2 (after 72 h). In general, the most abundant a phyla are Firmicutes and Bacteroidetes, while Bacilli and Bacteroidia dominate the b class level

Fig. 3

Gut microbiome diversity alterations in 13 ICU patients across two timepoints. a Principal coordinate analysis (PCoA) ordination plots of gut microbiome samples from patients at two timepoints (26 total samples). A significant shift in gut microbiome composition between the timepoints was identified by our PERMANOVA analysis (R² = 0.06, P = 0.004). Points represent samples from T1 (green) and T2 (orange). Gray solid lines connect samples from the same patient, with green and orange lines marking the centroids of T1 and T2 samples, respectively. Ellipses correspond to 95% confidence regions. b, c T2 samples showed significant reductions in both species-level Shannon Index (P = 0.048) and richness (P = 0.010) compared to T1. P-values were obtained using the two-sided Wilcoxon signed-rank test

Fig. 4

Microbial species from kefir exhibit different patterns of gut engraftment. T1 and T2 prevalences of the 12 microbial species found in kefir. There was an increased prevalence of three Lactobacillus species in ICU patients (Lactobacillus plantarum, Lactobacillus reuteri, and Lactobacillus rhamnosus). However, the prevalence of Lactobacillus acidophilus remained unchanged, and Bifidobacterium longum was detected less frequently at T2. Seven species were not detected in any of the analyzed stool samples

Fig. 5

Changes in gut microbiome wellness and GMWI species abundances in ICU patients. a GMWI increased significantly (P = 0.034, one-sided Wilcoxon signed-rank test) from T1 to T2. b The chart represents the fold change in relative abundances of GMWI species between T1 and T2, with the names of health-prevalent species depicted in blue (n = 1) and those of health-scarce species (n = 22) in brown. Notably, two health-scarce species, Hungatella hathewayi and Clostridium bolteae, exhibited a statistically significant decrease in relative abundance by T2 (P < 0.05, permutation test). The black horizontal lines indicate the extent of fold change, while the size of each circle reflects the prevalence of the respective GMWI species among the 26 stool samples gathered across both timepoints