Address for Correspondence Kentaro Shimizu, MD, Department of Clinical Quality Management, Osaka University Hospital, 2-15 Yamadaoka, Suita-City, Osaka 565-0871, Japan. Tel: +81 6 6879 5707; fax: +81 6 6879 5720; e-mail: email@example.com
Background The gut is an important target organ for injury after severe insult, and resolution of feeding intolerance is crucial for critically ill patients. We investigated gut flora and motility to evaluate the impact of gastrointestinal dysmotility on septic complications in patients with severe systemic inflammatory response syndrome (SIRS).
Methods Sixty-three ICU patients with severe SIRS were divided into two groups depending on their intestinal condition. Patients with feeding intolerance comprised patients who had feeding intolerance, defined as ≥300 mL reflux from nasal gastric feeding tube in 24 h, and patients without feeding intolerance comprised patients with no feeding intolerance. We compared fecal microflora, incidences of bacteremia, and mortality between these groups.
Key Results Analysis of feces showed that patients with feeding intolerance had significantly lower numbers of total obligate anaerobes including Bacteroidaceae and Bifidobacterium, higher numbers of Staphylococcus, lower concentrations of acetic acid and propionic acid, and higher concentrations of succinic acid and lactic acid than those in patients without feeding intolerance (P ≤ 0.05). Patients with feeding intolerance had higher incidences of bacteremia (86%vs 18%) and mortality (64%vs 20%) than did patients without feeding intolerance (P ≤ 0.05).
Conclusions & Inferences Gut flora and organic acids were significantly altered in patients with severe SIRS complicated by gastrointestinal dysmotility, which was associated with higher septic mortality in SIRS patients.
The gut is considered an important target organ for injury after severe insults such as sepsis, trauma, and burns. Gut motility in critically ill patients may be disturbed by many factors, including ischemia, analgesic drugs, adrenergic agents, fluid management, and pre-existing illnesses such as diabetes.1 This motor stasis leads to intolerance to enteral feeding, increased mucosal permeability for endoluminal mediators and bacteria, and the development of systemic inflammatory response syndrome (SIRS).2 Montejo reported that enteral nutrition-related gastrointestinal complications in critically ill patients were present in 251 (62.8%) of 400 patients and that the withdrawal of enteral nutrition occurred in 15.2%. Complications included high gastric residuals (39%), constipation (15.7%), diarrhea (14.7%), abdominal distention, vomiting, and regurgitation. Patients with gastrointestinal complications had significantly higher mortality than those without such complications.3 Treatment of impaired gastrointestinal motility in critically ill patients is difficult because the mechanisms underlying dysmotility are usually complex, the patients’ pre-existing characteristics are diverse, and the availability of pharmacological therapies is limited.4
Commensal gut flora has important metabolic, trophic, and protective functions.5 Gut flora is mainly influenced by host, diet, and intrabacterial factors. Host factors include gastric juice, bile acid, bowel movement, oxygen, sex, aging, antibiotics, and stress. Although recent reports indicate that gut flora could influence components of the enteric nervous system and motility in the gut, few clinical reports have addressed the relation between gut flora and gastrointestinal motility in critically ill patients. We recently evaluated microflora and changes in gastrointestinal environment in patients with severe SIRS and found significant deterioration.6
In the present study, we investigated gut flora and gastrointestinal motility to evaluate the impact of gastrointestinal dysmotility on septic complications and mortality in patients with severe SIRS.
Materials and methods
The present study enrolled 63 patients with severe SIRS who were admitted to the Department of Traumatology and Acute Critical Medicine, Osaka University Graduate School of Medicine, Japan, and treated in the ICU for more than 2 days, and who had not undergone any abdominal surgery. Severe SIRS was diagnosed in patients who fulfilled the criteria for SIRS of the American College of Chest Physicians and the Society of Critical Care Medicine7 and who had a serum C-reactive protein (CRP) level >10 mg dL−1. Enteral nutrition (Lifelon®-Q10; Nisshin Pharma Inc., Tokyo, Japan; protein 5 g, fat 3.4 g, and carbohydrate 12.5 g per 100 kcal, osmolarity 370 mOsm L−1) was initiated via duodenal feeding within 3 days of admission for all patients. Continuous enteral feeding was initiated from 500 kcal day−1 and gradually increased to more than 25 kcal kg−1 depending on the patient’s condition. We used a nasal gastric feeding tube to evaluate reflux every 8 h. When reflux from the nasal gastric feeding tube was greater than 300 mL day−1, 10 mg of intravenous metoclopramide was administered twice daily.8 If infectious complications occurred, antibiotics were chosen based on the underlying clinical syndrome and the results of microbiologic cultures and gram staining. The strategy for the use of antibiotics was identical throughout the study period. Fecal samples were collected serially after admission and analyzed. This study was approved by the Institutional Review Board of Osaka University, and informed consent was obtained from the family of each patient.
Severe SIRS patients were divided into two groups depending on intestinal condition. The patients without feeding intolerance comprised patients who had no feeding intolerance during their stay in the ICU. The patients with feeding intolerance comprised patients who had feeding intolerance due to intestinal dysmotility, which was defined as more than 300 mL of reflux from the nasal gastric feeding tube in 24 h.8 Gastrointestinal dysmotility was diagnosed on the day the patient met this criterion.
Fecal bacteriologic culture
Feces were collected in a test tube, which was maintained under anaerobic conditions with CO2 saturation. The test tube was cooled in an icebox before culture. VL-G roll tube agar9 supplemented with 0.2% cellobiose and 0.2% maltose (modified VL-G roll tube agar) was used to determine total anaerobe counts. Different media were used for selective isolation of different microorganisms: modified VL-G roll tube agar to which 80 μg mL−1 vancomycin and 1 μg mL−1 kanamycin were added for Bacteroidaceae; CW agar (Nikken Bio Medical Laboratory Inc., Kyoto, Japan) for lecithinase-positive Clostridium; MPN roll tube agar10 for Bifidobacterium; COBA agar11 for Enterococcus; LBS agar (Becton Dickinson and Company, Cockeysville, MD, USA) supplemented with 0.8% Laboratory Lemco powder (Oxoid Co. Ltd., Basingstoke, UK) for Lactobacillus; Staphylococcus medium no. 110 agar (Nissui Pharmaceutical Co. Ltd., Tokyo, Japan) for Staphylococcus; DHL agar (Nissui) for Enterobacteriaceae; NAC agar (Nissui) for Pseudomonas; and GS agar (Nissui) for Candida. Reproducibility and stability of these measurements were shown previously.12,13
Determination of fecal organic acid concentrations
A portion of the feces was isolated, weighed, mixed with 0.15 mol L−1 perchloric acid at a 4-fold volume, and stored at 4 °C for 12 h. The mixture was centrifuged at 4 °C at 20 400 g for 10 min. The supernatant was filtered with a 0.45-μm membrane filter (Millipore Japan Ltd., Tokyo, Japan) and sterilized. The sample was analyzed for organic acids by high-performance liquid chromatography performed with a Waters system (Waters 432 Conductivity Detector; Waters Co., Milford, MA, USA) equipped with two columns (Shodex RSpack KC-811; Showa Denko Co. Ltd., Tokyo, Japan). The concentrations of organic acids were calculated with the use of external standards, and the reproducibility and stability of these measurements have been previously shown.14
Surveillance and definition of infection
Bacterial infection was diagnosed in accordance with the Centers for Disease Control definitions.15 Body temperature was measured continuously. Surveillance cultures from blood and sputum were performed routinely for each patient. In cases of suspected infection, laboratory testing, chest X-ray, and computed tomography scanning were performed as necessary. Bacteremia was defined as a positive blood culture.
Our previous results showed that the decrease of total obligate anaerobes was significantly associated with septic mortality in patients with severe SIRS and was the most significant prognostic factor in gut flora.16,17 To evaluate the impact of gastrointestinal dysmotility on gut flora and gastrointestinal environment in patients with severe SIRS, we compared the results of fecal analysis between the patients without feeding intolerance, when the number of total obligate anaerobes was at the minimum, and the patients with feeding intolerance, when the first fecal sample was obtained after the diagnosis of intestinal dysmotility. Results are expressed as mean ± SD values. Statistical analysis was performed with Fisher’s exact probability test and an unpaired Student’s t-test and Mann–Whitney test where appropriate. We also performed logistic regression analysis using mortality as the binary outcome variable, and age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II score on admission, number of antibiotic types, duration of antibiotic usage, and intestinal dysmotility as potential predictor covariates. Statistical significance was determined at P < 0.05.
The study cohort comprised 33 men and 30 women with a mean (±SD) age of 58.8 ± 20.7 years. Severe SIRS was caused by sepsis in 47 patients, trauma in 12 patients, and major burns in four patients. Patient characteristics are listed in Table 1. There were 49 patients without feeding intolerance and 14 patients with feeding intolerance. The two groups did not differ significantly in terms of age, sex, and APACHE II score on admission. Catecholamines were used in nine of the 14 patients with feeding intolerance and in four of the 49 patients without feeding intolerance. There was a significant difference in the rate of catecholamine use between the two groups (P <0.05). Sepsis was the main cause of SIRS in the patients with feeding intolerance, and seven of the 14 patients had abdominal inflammation such as peritonitis or enteritis, with the principal bacteria detected being Staphylococcus species in four patients, and Pseudomonas aeruginosa, Escherichia coli, and Candida albicans in one of each of the remaining three patients.
Table 1. Characteristics of patients with severe systemic inflammatory response syndrome
Patients without feeding intolerance
Patients with feeding intolerance
APACHE, Acute Physiology and Chronic Health Evaluation; SIRS, systemic inflammatory response syndrome.
*P <0.05 compared with patients without feeding intolerance.
Age (mean ± SD)
62.6 ± 19.3
59.4 ± 15.5
APACHE II score on admission (mean ± SD)
18.8 ± 13.7
17.1 ± 8.9
No. antibiotic types*
3.9 ± 3.2
6.6 ± 2.7
Duration of antibiotics usage* (days)
15.9 ± 15.7
48.0 ± 30.3
Origin of SIRS
In the patients with feeding intolerance, intestinal dysmotility was diagnosed at 14.4 ± 14.9 days after admission (mean ± SD). The mean value of maximum reflux from the nasal gastric tube was 990 ± 700 mL day−1 (mean ± SD) at 6.8 ± 9.4 days after the diagnosis of intestinal dysmotility.
Analysis of fecal flora is shown in Table 2. The sample was obtained 5.7 ± 4.1 days after admission in the patients without feeding intolerance and 25.1 ± 21.0 days after admission in the patients with feeding intolerance (mean ± SD). The numbers of total obligate anaerobes including Bacteroidaceae and Bifidobacterium in the patients with feeding intolerance were significantly lower than those in the patients without feeding intolerance (P < 0.05). The numbers of Staphylococcus species in the patients with feeding intolerance were significantly higher than those in the patients without feeding intolerance (P < 0.05). Analysis of fecal organic acid concentrations is shown in Fig. 1. The concentrations of acetic acid and propionic acid in the patients with feeding intolerance were significantly lower than those in the patients without feeding intolerance (P <0.05), but the concentrations of succinic acid and lactic acid in the patients with feeding intolerance were significantly higher than those in the patients without feeding intolerance (P <0.05).
Table 2. Fecal gut flora in patients with and without feeding intolerance
Patients without feeding intolerance
Patients with feeding intolerance
ND, not detected.
*P <0.05 compared with patients without feeding intolerance. Values are mean ± SD (Log10 colony-forming unit g−1 feces).
Total obligate anaerobes
9.4 ± 1.3
8.2 ± 2.7*
10.5 ± 0.5
8.7 ± 1.9
7.0 ± 3.4*
10.1 ± 0.4
7.4 ± 2.7
5.6 ± 3.3*
9.6 ± 0.7
2.5 ± 1.3
2.8 ± 2.2
2.1 ± 0.7
Total facultative anaerobes
8.4 ± 1.4
7.6 ± 1.4
7.5 ± 0.4
4.9 ± 2.5
4.5 ± 2.4
5.0 ± 1.0
6.0 ± 2.3
3.7 ± 2.8*
7.4 ± 0.8
7.4 ± 2.1
6.6 ± 2.4
7.0 ± 0.9
3.9 ± 1.9
5.3 ± 1.6*
2.7 ± 0.8
2.8 ± 1.7
2.9 ± 1.9
2.9 ± 1.5
3.5 ± 2.1
2.0 ± 0.5
The incidence of bacteremia in the patients with feeding intolerance (86%) was significantly higher than that in the patients without feeding intolerance (18%) (P < 0.05). Mortality due to septic multiple organ dysfunction syndrome (MODS) in the patients with feeding intolerance (64%) was significantly higher than that in the patients without feeding intolerance (20%) (P <0.05). Logistic regression analysis revealed that dysmotility was a significant prognostic factor for mortality and bacteremia (Table 3).
Table 3. Results of multivariate logistic regression analysis
Coeff(β), coefficient; SE(β), standard error of coefficient; OR, odds ratio; 95% CI, 95% confidence interval; APACHE, Acute Physiology and Chronic Health Evaluation.
APACHE II score on admission
Duration of antibiotics usage
No. antibiotic types
Blood culture analysis in the patients without feeding intolerance detected Staphylococcus species in four patients (Staphylococcus epidermidis in one) and Escherichia coli in two patients. Seven different species of bacteria, one per patient, were detected in seven patients. Blood culture analysis in the patients with feeding intolerance detected Staphylococcus species in seven patients (S. epidermidis in two patients), Pseudomonas aeruginosa in three patients, E. coli in three patients, Enterococcus faecalis in two patients, and four other species of bacteria in one patient.
We previously evaluated microflora and changes in the gastrointestinal environment in patients with severe SIRS and showed significant deterioration.6 Analysis of fecal flora confirmed that SIRS patients had markedly lower total anaerobic bacterial counts, especially those of Bifidobacterium and Lactobacillus, and higher counts of Staphylococcus and Pseudomonas group bacteria than did healthy volunteers.
The present results clearly show statistically significant alteration of gut flora in patients with SIRS complicated by gastrointestinal dysmotility when compared to the patients without dysmotility. The number of total obligate anaerobes in the patients with feeding intolerance was significantly lower than that in the patients without feeding intolerance, and beneficial bacteria, including Bifidobacterium, were markedly decreased in the patients with feeding intolerance. In contrast, the number of Staphylococcus in the patients with feeding intolerance was significantly higher than that in the patients without feeding intolerance. These results indicate that gastrointestinal dysmotility in critically ill patients was associated with alteration of gut flora. Although the timing of fecal sampling for analysis was different between the two groups, our results showed that fecal flora in the patients with gastrointestinal dysmotility was further altered as compared with the most altered flora in the patients without feeding intolerance.
Gut motility is regulated by the enteric nervous system, interstitial cells of Cajal, and gastrointestinal smooth muscle. In critically ill patients, gastrointestinal motility is sensitive to any kind of stress such as abdominal surgery, trauma, SIRS, sepsis, hypoperfusion and hypoxemia, use of catecholamines, and imbalances in pH, glucose and electrolytes, and fluid status.18 In the present study, half of the patients with feeding intolerance had peritonitis or enteritis. Abdominal inflammation may be a cause for gastrointestinal dysmotility. Because there was a significant difference in the use of catecholamines in the two groups, catecholamines may also have affected gastrointestinal motility in our study. The influence of catecholamines on gut flora will require further elucidation.
Several studies have suggested that gut flora can influence gastrointestinal motility. In animal studies, electrophysiologic examinations of germ-free and conventionally raised animals indicated that in the absence of intestinal microbiota, intestinal motility is disrupted.19 Hooper et al. colonized germ-free mice with Bacteroides thetaiotaomicron and compared ileal gene expression in germ-free and colonized mice by DNA microarrays.20 Their results showed that there was not only increased expression of genes involved in nutrient absorption and mucosal barrier fortification but also an increased number of motility-related genes such as that of gamma-aminobutyric acid. In clinical studies, irritable bowel syndrome patients who have a significant gastrointestinal disorder have significantly altered fecal microbiota compared with healthy controls by 16S ribosomal RNA gene cloning.21 These results suggest that gut flora can be associated with gastrointestinal motility,22 and that alteration of gut flora in SIRS patients could exert influence on gut motility.
The gut is a target organ for injury after severe insult,23 and dysmotility can affect the intestinal microflora. Runkel et al. reported that in a rat model, the bacterial counts in the gut increased significantly after subcutaneous infusion of morphine.24 In the present study, alteration of gut flora was greater in the patients with feeding intolerance than in those without feeding intolerance. Bacterial overgrowth as a consequence of gut dysmotility may lead to bacterial translocation and bacteremia in critically ill patients. Thus, gastrointestinal dysmotility with altered gut flora as observed in our study can be both a cause and a consequence of critical illness. Further study is required to evaluate the relation between gastrointestinal dysmotility and critical illness.
In our study, the analysis of fecal organic acid concentrations showed that short-chain fatty acids (SCFAs) such as acetate, propionate, and butyrate were lower in patients with feeding intolerance than in patients without feeding intolerance. Short-chain fatty acids are the metabolic end products of colonic microflora fermentation of carbohydrates and glycoproteins.25 Short-chain fatty acids are utilized mainly by intestinal epithelial cells as energy substrates, and some are absorbed into the portal flow to the liver and utilized as systemic energy sources.26 Short-chain fatty acids have effects on gut motility, especially in stimulating ileal peristaltic contractions and tonic activity. Kamath et al. reported that in 18 healthy human volunteers, ileal motility was stimulated more often by SCFAs than by similar volumes of air or saline.27 The mechanisms may involve a direct interaction with intrinsic and extrinsic nerves and ileal and colonic smooth muscle. Such results, together with our findings, indicate that gastrointestinal dysmotility in critically ill patients may be associated with decreased SCFAs in the gut.
Our results also showed that the concentrations of succinic acid and lactic acid in the patients with feeding intolerance were significantly higher than those in the patients without feeding intolerance. Sakata et al. reported that accumulation of lactic acid in the gut leads to lactic acidosis and that both low pH and succinic acid reduce motility.28 These results indicate that not only a lack of SCFAs but also an accumulation of succinic acid and lactic acid in the patients with feeding intolerance may be associated with gastrointestinal dysmotility.
Gut dysmotility was significantly associated with septic mortality in patients with severe SIRS. Because the APACHE II scores on admission were similar in the two groups, our results showed that the severity of SIRS was enhanced in the patients with gastrointestinal dysmotility. Mortality due to septic MODS in the patients with feeding intolerance (64%) was significantly higher than that in the patients without feeding intolerance (20%), indicating that the patients with severe SIRS and gastrointestinal dysmotility have altered gut flora that would lead to an ‘undrained abscess’. This condition may cause bacterial translocation and may be associated with bacteremia and mortality. Further careful analysis is required to evaluate bacterial translocation in such patients.
The results of the present study suggest that maintenance of gut flora and motility may be required for critically ill patients. We recently reported in a preliminary study that synbiotics maintain the gut flora and gastrointestinal environment and decrease the incidence of septic complications in patients with severe SIRS.29 Prokinetic drugs are also required for gut motility. The effects of metoclopramide and erythromycin have been reported, but the rapid development of tachyphylaxis within the first 3 days of therapy was a problem.8 Further studies are needed to develop an effective treatment strategy in critically ill patients with gastrointestinal dysmotility.
In conclusion, we found that the gut flora and gastrointestinal environment were significantly altered in patients with severe SIRS complicated by gastrointestinal dysmotility. Gastrointestinal dysmotility was associated with higher septic mortality in SIRS patients.
This work was supported by a grant from the Ministry for Education, Science, and Culture of Japan.
This work was performed at the Department of Traumatology and Acute Critical Medicine, Osaka University Graduate School of Medicine, 2-15 Yamadaoka, Suita-City, Osaka 565-0871, Japan.
KS was the primary author of the manuscript. HO made substantial contributions to analysis and interpretation of data and gave final approval of the study. TA, KN, and MM are professional microbiological researchers from Yakult Central Institute for Microbiological Research; they cultured and checked the gut flora quantitatively. OT was involved in statistical analysis. MG, AO, SY, and YN collected feces from each patient. AM made a contribution to data analysis and interpretation of data. HS and YK were the general supervisors of this research from its inception.