Severe pneumonia is a life-threatening lower respiratory tract infection characterized by rapid clinical deterioration and high mortality. In addition to respiratory failure, severe pneumonia may result in multi-organ dysfunction, including acute kidney injury, which significantly worsens prognosis. During severe inflammation, the composition of the gut microbiota undergoes significant alterations, which can disrupt microbial balance. This dysbiosis is associated with exacerbated local and systemic inflammation, increased uremic toxin concentrations, elevated intestinal permeability, and the development of endotoxemia and immune deficiencies (1). Probiotics have been shown to enhance intestinal barrier function, reduce the host pathogen load, competitively inhibit pathogenic bacteria, regulate host inflammatory responses (2), and modulate the immune system (3). Some studies suggest that probiotics may reduce the risk of sepsis and lower respiratory tract infections (4). However, researchers conducted a randomized controlled trial and found that probiotic administration did not reduce the incidence of ventilator-associated pneumonia (VAP), duration of antibiotic use, or length of ICU stay in critically ill patients (1). They believe that although critically ill patients exhibit a reduction in commensal microbiota, overgrowth of potential pathogens, and consequently, a highly perturbed microbial community (5), probiotics may not improve clinically significant outcomes associated with dysbiosis in such circumstances. Similarly, recent research has demonstrated that, in mechanically ventilated patients with severe conditions, probiotics did not reduce the incidence of ventilator-associated pneumonia compared to a placebo (1). Despite numerous studies exploring the effects of probiotics on various health outcomes, results remain conflicting. These discrepancies may arise from several factors, including differences in patient populations, probiotic strains used, and variations in study designs.

Currently, the management of acute kidney injury (AKI) caused by sepsis is primarily limited to treating the underlying cause (6), and in severe cases, addressing its complications through kidney replacement therapy (KRT) (7). Therefore, there is an urgent need to explore alternative therapeutic approaches for AKI. In 2023, researchers conducted a multicenter randomized controlled trial to investigate whether probiotics could promote renal function recovery during AKI in sepsis. The results demonstrated that administering probiotics for seven consecutive days was safe but did not increase the likelihood of renal function recovery among sepsis patients (8).

As AKI is a syndrome that elicits a strong systemic inflammatory response (9), it has been demonstrated that attenuating this response can improve renal function and mitigate parenchymal damage (10). Animal experiments have revealed a bidirectional relationship between host AKI and gut microbiota. The gut microbiota can significantly influence the prognosis of AKI, while AKI can also induce alterations in the gut microbiota. This interaction is termed the “gut-kidney axis.” The bidirectional relationship between the gut and kidney underscores the critical interplay between these two organs, with significant clinical implications. This inter-organ communication suggests potential therapeutic avenues via microbiota modulation. Modulating the microbiota through probiotic administration to ameliorate gut dysbiosis during AKI appears to improve the prognosis of patients with these conditions (11). Current research primarily focuses on whether probiotics can improve the prognosis of patients with severe pneumonia and whether they can promote the recovery of AKI after its onset. Despite growing interest in the gut–kidney axis, there is limited evidence on whether probiotic administration can prevent AKI in patients with severe pneumonia who have no prior renal disease. This retrospective study investigates whether probiotics use is associated with a reduced incidence of AKI in this patient population.

Based on the inclusion and exclusion criteria, a total of 267 cases were included, with 190 patients (71.16%) not developing AKI and 77 patients (28.84%) developing AKI. The AKI group had lower male proportion (12.99% vs. 26.31%, p = 0.018), higher CRP (120.1 mg/L vs. 74.26 mg/L, p = 0.0076), higher PCT (1.22 ng/mL vs. 0.51 ng/mL, p = 0.001), lower PLT (160 vs. 185, p = 0.0089), higher SOFA score (10 vs. 7, p = 0.001), lower ALB (29.8 g/L vs. 32.9 g/L, p = 0.008), higher Cr (111 vs. 67, p = 0.001), and higher BUN (11.21 vs. 7.845, p = 0.001). No significant differences were found in age, D - dimer, WBC, OI, or the rates of hypertension, diabetes, and liver dysfunction between the two groups (p > 0.05) (Table 1).

Probiotics use was higher in the non - AKI group (51.05% vs. 31.17%, p = 0.003).

According to the statistical results, 121 patients received probiotics, among whom 24 patients (19.83%) developed AKI; 146 patients did not receive probiotics, among whom 53 patients (36.30%) developed AKI. The chi-square test results showed that the proportion of AKI occurrence was statistically significant between the two groups before adjustment (χ² = 8.74, p = 0.003). Using the occurrence of AKI as the dependent variable, multivariate logistic analysis was performed incorporating indicators (gender, CRP, PCT, PLT, SOFA, ALB, Cr, and BUN) with statistical differences between the two groups.

As shown in Table 2, after adjusting for the aforementioned (gender, CRP, PCT, PLT, SOFA, ALB, Cr, and BUN) factors, gender, CRP, PCT, PLT, BUN and Sofa score showed no significant correlation with AKI occurrence (all p > 0.05).

Cr levels were positively correlated with AKI occurrence (OR = 1.020, 95% CI = 1.011–1.029, p = 0.000). ALB levels were inversely correlated with AKI occurrence (OR = 0.904, 95% CI = 0.842–0.970, p = 0.005).

Probiotics use was associated with reduced AKI occurrence (OR = 0.483, 95% CI = 0.249–0.935, p = 0.031). Patients using probiotics had about half the odds of developing AKI compared to non - users.

According to the KDIGO criteria, AKI is classified into stages 1, 2, and 3. The results of the Ordered Logistic Regression (Table 3) indicate that Probiotic use was negatively associated with AKI stage (Coefficient = −0.775, Std. err. = 0.282, z = −2.74, p = 0.006). The 95% CI was −1.328 to −0.221, indicating that probiotics use was significantly related to lower AKI stages.

The Brant test was conducted to assess the proportional odds assumption of the ordered logistic regression model (X² = 2.18, p = 0.3358), thus the ordered logistic regression model is appropriate for the data.

The log-likelihood of the intercept-only model is −235.709, while the log-likelihood of the full model is −231.770. The deviance (D) with 263 degrees of freedom is 463.539. The likelihood ratio test statistic (LR) with 1 degree of freedom is 7.879, with a p-value of 0.005, indicating the model is statistically significant compared to the intercept-only model.

Using the occurrence of AKI stages as the dependent variable, multivariate logistic analysis was performed incorporating statistically significant indicators from the two groups mentioned above. The results (Table 4) showed that probiotics could reduce the incidence of stage 1 AKI (OR: 0.4023, 95% CI = 0.1895–0.8540, p = 0.018) but not stage 2 (OR: 0.683, 95% CI = 0.2083–2.242, p = 0.53) or stage 3 AKI (OR: 0.9007, 95% CI = 0.2851–2.845, p = 0.859).

Based on the highest AKI stage during the patient’s ICU stay, there were 44 patients with stage 1 AKI, 15 with stage 2 AKI, and 18 with stage 3 AKI. Before matching, there was a statistically significant difference in the incidence of stage 1 AKI between the two groups (p = 0.009); however, there were no statistically significant differences in the incidence of stage 2 AKI (p = 0.337) or stage 3 AKI (p = 0.570). We performed a univariate analysis of all potential covariates and found that only creatinine was significantly different between the probiotic and non-probiotic groups (p = 0.024). After propensity score matching with creatinine as a covariate, a total of 224 cases were matched (112 in the probiotic group and 112 in the non-probiotic group). Graph of PSM was shown in Figure 1. After matching, the two groups achieved good balance in baseline characteristics (Table 5). The occurrence of stage 1 AKI (p = 0.019) remained statistically significant. There were no statistically significant differences between the two groups for stage 2 AKI (p = 0.757) or stage 3 AKI (p = 0.757).

To demonstrate balance after PSM, we calculated the standardized mean difference (SMD) for creatinine levels between the matched groups. The SMD was found to be −0.00286, this ensures that the matched groups are comparable, thereby strengthening the validity of our findings.

Among mechanically ventilated patients with severe pneumonia, we found that probiotics can reduce the incidence of stage 1 AKI. Our findings add new evidence to the clinical application of probiotics. AKI presents significant challenges in pediatric care (15). Thus future similar perspective studies could be also made in children.

Common causes of AKI in patients with severe pneumonia include drug-induced nephrotoxicity (16), extra-renal tissue damage (17), and ischemia–reperfusion injury (IRI) (18). Studies have shown that oxidative stress (19), inflammation [unreferenced], mitochondrial damage (20), and immune dysfunction (21) are core mechanisms leading to the development and progression of AKI.

Recent studies have demonstrated that probiotics can mitigate the severity of gastrointestinal adverse reactions induced by antitumor treatment and restore disrupted gut microbiota (22). In a definitive efficacy trial, probiotics were employed as an adjunctive therapy for patients with depression (23). Uremic toxins derived from gut dysbiosis can exacerbate renal dysfunction and fibrosis. For instance, the gut-derived metabolite trimethylamine-N-oxide (TMAO) directly contributes to the progression of renal dysfunction and is associated with an increased risk of mortality (24). Certain probiotics prevent AKI by reducing inflammation and lowering oxidative stress levels. Lactobacillus salivarius BP121 exerts renoprotective effects primarily by modulating tight junction protein assembly through AMPK and TLR4 pathways and inhibiting the production of uremic toxins such as indoxyl sulfate (IS) and p-cresyl sulfate (PCS) (25). In mice with Escherichia coli-induced pyelonephritis, probiotic administration before and after sepsis significantly improved renal function, reduced inflammation, and attenuated renal fibrosis (26). Probiotics competitively inhibit pathogenic bacteria, maintain gut microbiota balance, and reduce the translocation of toxins and inflammatory mediators into the bloodstream. Additionally, they enhance gut epithelial barrier function, decrease endotoxin and bacterial translocation, and attenuate systemic inflammatory responses.

After AKI onset, probiotics have demonstrated efficacy in improving renal function, potentially by lowering creatinine and urea levels and significantly reducing pro-inflammatory cytokines like IL-1β, IL-6, TNF-α, and MCP-1 (27). They modulate immune cell activity, curb the release of pro-inflammatory cytokines (e.g., TNF-α, IL-6), and mitigate systemic inflammatory responses, while also boosting anti-inflammatory cytokines (e.g., IL-10) to alleviate inflammatory kidney damage. Probiotics enhance immune function and infection resistance by activating immune cells such as macrophages and T-cells. Gut dysbiosis can heighten uremic toxin production and foster AKI development (27). In a mouse renal ischemia–reperfusion injury (IRI) model (28), the probiotic Bifidobacterium bifidum (BGN4) increased gut microbiome evenness and curbed the proliferation of AKI-linked bacteria like Enterobacteriaceae and Bacteroidaceae. BGN4 intervention notably reduced tubular damage scores, neutrophil and macrophage infiltration, and renal IL-6 mRNA expression in mice (28). Additionally, BGN4 lowered plasma low-density lipoprotein, alanine aminotransferase, and aspartate aminotransferase levels, indicating its potential to alleviate renal IRI and AKI-related liver damage (29).

Additionally, some probiotics, such as ATCC 4356, can activate the nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase-1 (HO1) axis, a crucial pathway for antioxidant stress, reduce the expression of malondialdehyde (MDA), and promote the expression of superoxide dismutase (SOD) and Nrf2. ATCC 4356 also exerts anti-inflammatory effects by regulating the PI3K/Akt signaling pathway (30).

Complement system is considered a key pathogenic mediator in the development of several renal diseases (31). Kidney is particularly susceptible to complement-mediated injury mainly due to its ultrafiltration function, low expression of complement regulatory factors, and local complement production (32). It can cause kidney injury and also cause AKI. Higher inflammation can lead to a hyperactive complement system, which can damage the kidney. Probiotics may influence the activation and regulation of complement components. This helps to reduce the release of factors that trigger complement activation.

Our study found that probiotics reduced the incidence of stage 1 acute kidney injury (AKI) but were ineffective for stage 2 and stage 3 AKI. This may be attributed to the fact that stage 1 AKI is milder, allowing probiotics to effectively intervene by modulating the gut microbiota and reducing inflammation. In contrast, stage 2 and stage 3 AKI are more severe and involve more complex pathological mechanisms, making single-intervention strategies such as probiotics insufficient. Probiotics primarily exert their effects by modulating the gut microbiota, reducing inflammation, and mitigating oxidative stress, which may be particularly effective in the early stages of AKI. In late-stage AKI, however, widespread systemic inflammation and multi-organ damage occur, and the effects of probiotics may be inadequate to reverse these complex pathological changes. Additionally, in stage 1 AKI, gut barrier function may remain relatively intact, enabling probiotics to exert their beneficial effects. In stage 2 and stage 3 AKI, severe damage to the gut barrier may hinder the ability of probiotics to function effectively. The immune response in stage 1 AKI is also milder, and probiotics may alleviate renal damage by modulating the immune system. In contrast, the immune response in late-stage AKI is more intense, and the regulatory effects of probiotics may be insufficient to control excessive immune activation. Moreover, the dosage and duration of probiotic administration may be adequate for stage 1 AKI but may require higher doses or extended treatment periods to address the severity of stage 2 and stage 3 AKI. Furthermore, patients with stage 2 and stage 3 AKI often present with additional complications or receive multiple concurrent treatments, which may interfere with the efficacy of probiotics.

Severe pneumonia and sepsis are closely related and interconnected. Pneumonia, especially severe pneumonia, is one of the most common causes of sepsis. The pathogenetic mechanisms of both diseases are similar in that they both involve the body’s immune response to infection, which can result in tissue damage and organ dysfunction. Additionally, sepsis patients are also at an increased risk of developing pneumonia, due to their weakened immune systems and the use of invasive medical devices. In clinical practice, a significant percentage of pneumonia cases are misdiagnosed as septic shock, and severe sepsis is often identified as a critical complication of pneumonia. Therefore, given the close relationship between severe pneumonia and sepsis, we shift our focus from severe pneumonia to sepsis to further explore the relevant research questions.

Probiotics were associated with a lower incidence of stage 1 AKI in patients with severe pneumonia. However, our research was a single-center retrospective study, and multi-center randomized controlled studies would be needed in the future for validation. Defining optimal probiotic strains, dosages, and treatment windows would enhance the translational impact of the conclusions in future studies.