Conceived and designed the experiments: KvW KL LvL WP WB CD. Performed the experiments: KvW. Analyzed the data: KvW KL WB CD. Contributed reagents/materials/analysis tools: LvL WP. Wrote the paper: KvW KL WB CD. Critical revision manuscript: LvL WP.
The authors have declared that no competing interests exist.
Splanchnic hypoperfusion is common in various pathophysiological conditions and often considered to lead to gut dysfunction. While it is known that physiological situations such as physical exercise also result in splanchnic hypoperfusion, the consequences of flow redistribution at the expense of abdominal organs remained to be determined. This study focuses on the effects of splanchnic hypoperfusion on the gut, and the relationship between hypoperfusion, intestinal injury and permeability during physical exercise in healthy men.
Healthy men cycled for 60 minutes at 70% of maximum workload capacity. Splanchnic hypoperfusion was assessed using gastric tonometry. Blood, sampled every 10 minutes, was analyzed for enterocyte damage parameters (intestinal fatty acid binding protein (I-FABP) and ileal bile acid binding protein (I-BABP)). Changes in intestinal permeability were assessed using sugar probes. Furthermore, liver and renal parameters were assessed. Splanchnic perfusion rapidly decreased during exercise, reflected by increased gapg-apCO2 from −0.85±0.15 to 0.85±0.42 kPa (p<0.001). Hypoperfusion increased plasma I-FABP (615±118 vs. 309±46 pg/ml, p<0.001) and I-BABP (14.30±2.20 vs. 5.06±1.27 ng/ml, p<0.001), and hypoperfusion correlated significantly with this small intestinal damage (rS = 0.59; p<0.001). Last of all, plasma analysis revealed an increase in small intestinal permeability after exercise (p<0.001), which correlated with intestinal injury (rS = 0.50; p<0.001). Liver parameters, but not renal parameters were elevated.
Exercise-induced splanchnic hypoperfusion results in quantifiable small intestinal injury. Importantly, the extent of intestinal injury correlates with transiently increased small intestinal permeability, indicating gut barrier dysfunction in healthy individuals. These physiological observations increase our knowledge of splanchnic hypoperfusion sequelae, and may help to understand and prevent these phenomena in patients.
Splanchnic hypoperfusion occurs in many pathophysiological conditions, and is often considered to lead to epithelial barrier dysfunction. Vascular disease, trauma, and shock can induce splanchnic hypoperfusion and ischemia
Splanchnic hypoperfusion also occurs in physiological conditions. Young, healthy individuals endure episodes of splanchnic hypoperfusion during strenuous physical exercise
In the current study, we focus on the direct effects of splanchnic hypoperfusion on gut, liver, and kidney epithelium in healthy young volunteers who perform moderate-to-high intensity physical exercise. This study is based on the fact that functional splanchnic hypoperfusion occurs during physical exercise, when rapid redistribution of the splanchnic blood flow occurs to secure supply of adequate amounts of oxygen and energy to the active muscle tissue, heart, and lungs
In the current study, we determined the consequences of exercise-induced splanchnic hypoperfusion on gut, liver and kidney epithelium, with particular emphasis on intestinal barrier integrity loss in healthy individuals.
Tonometry revealed a significant increase of gastric-arterialized pCO2 (gapg-apCO2) during exercise, from −0.85±0.15 kPa to 0.85±0.42 kPa (p<0.001) at completion of exercise bout, indicating functional splanchnic hypoperfusion (
To assess hypoperfusion-induced effects on the small intestinal epithelium, plasma intestinal fatty acid binding protein (I-FABP) levels were determined, which increased rapidly, from 309±46 pg/ml at baseline to 615±118 pg/ml (p<0.001) after cycling (
The short half-life of I-FABP in plasma (approximately 11 minutes) allows to analyze the correlation of enterocyte damage with splanchnic hypoperfusion
Similar to I-FABP, cycling increased ileal bile acid binding protein (I-BABP) levels (5.06±1.27 to 14.3±2.20 ng/ml (p<0.001;
Two individuals reported nausea during the GI perfusion measurements with the nasogastric tube
To obtain information on liver injury during exercise, we analyzed a set of four parameters: liver fatty acid binding protein (L-FABP), alanine transaminase (ALT), aspartate transaminase (AST), and alpha-glutathione S-transferase (alpha-GST). Plasma L-FABP levels significantly increased in all individuals upon exercise (4.75±0.30 to 14.2±1.38 ng/ml (p<0.001;
Plasma L-FABP (
While we assumed that exercise caused renal damage, no significant changes in urinary N-acetyl-beta-(D)-glucosaminidase (NAG) levels were observed (p>0.05;
To analyze whether the exercise-induced GI hypoperfusion and consequent enterocyte damage resulted in GI permeability changes, we analyzed small intestinal permeability, which was assessed as lactulose/L-rhamnose (L/R) ratio in urine and plasma. All subjects were able to urinate within 10 minutes of the instructed time point, without using Foley catheters. Overall, a trend towards higher small intestinal permeability after exercise was observed in urine, which was especially pronounced in the first two hours of urinary collection (
Next, correlations between the 0–1 h urinary L/R ratio and the AUC of the plasma L/R ratios determined in the 0–1 h time period were evaluated for similarity, resulting in a RS of 0.42 (p<0.05; data not shown). Finally, the sucralose/erythritol (S/E) ratio was determined in the 5–24 h urinary collections to study large intestinal permeability changes. No changes in large intestinal permeability were observed (
To assess the translocation of bacterial endotoxin from the intestine into the circulation, EndoCab analysis was performed. The data did not reveal significant changes in endotoxin core antibody levels before and after cycling in the subgroup of participants that underwent permeability analysis (
Myeloperoxidase (MPO) and calprotectin were analyzed as parameters of innate inflammation. Plasma MPO levels increased significantly in all individuals upon exercise, from baseline 4.4±0.8 to 33±4.7 ng/ml after cycling (p<0.001;
The redistribution of blood away from the splanchnic area is an adaptation of the body enabling perfusion of critical tissues in both physiological and pathophysiological situations. During physical exercise, reduction of the splanchnic circulation occurs to meet the increased demand for oxygen and substrates in the active muscle, thereby inducing a physiological state of hypoperfusion in the GI system
In this study, we demonstrate that in healthy individuals, one hour of physical exercise results in splanchnic hypoperfusion and small intestinal damage, reflected by an increase in plasma I-FABP. Interestingly, the extent of intestinal damage significantly correlated with the exercise-induced hypoperfusion and with mild permeability changes in the small intestine. The most pronounced change in splanchnic perfusion, i.e. the most obvious increase in gapg-apCO2, occurred within the first 10 minutes of cycling, which implies that splanchnic perfusion rapidly decreases following the onset of exercise. The speed of adaptation of the splanchnic blood flow is further demonstrated by its fast recovery after cycling, which was also most prominent during the first 10 minutes following cessation of exercise. The rapid decline in splanchnic perfusion during cycling was followed by the development of significant intestinal damage with a time lag of 20 minutes. The data show a correlation between the exercise-induced splanchnic hypoperfusion and enterocyte damage (measured 20 minutes later), which is in line with animal studies and studies performed in critically ill patients demonstrating that intestinal hypoperfusion and ischemia result in intestinal cell damage
Having obtained strong indications for reduced splanchnic blood flow, we hypothesized that splanchnic outflow may be reduced during strenuous physical exercise, resulting in hepatocyte damage. Plasma levels of all assessed liver parameters (i.e. L-FABP, ALT, AST, and alpha-GST) elevated significantly after exercise. However, two other factors could be responsible for the phenomenon. First, the increase in these plasma parameters could theoretically be partly attributed to an exercise-induced reduction in plasma volume. The increases, however, largely exceeded the 10% decrease in plasma volume that has been described in athletes in more extreme settings than ours
In contrast to the intestinal and liver data, we did not observe indications for renal damage during exercise. At this stage, it is unclear whether there was no hypoperfusion in the kidneys during cycling, or whether the kidney is more resistant to physiological hypoperfusion. Previous studies support the former, showing relatively unchanged renal flow after short-term splanchnic blood redistribution
In addition to the observed epithelial cell integrity loss, activation of the inflammatory response was shown. Inflammation is characterized by neutrophil influx and activation, resulting in the release of MPO and calprotectin, as seen in the present study. MPO plays a role in oxidative processes, while calprotectin and one of its constituents, myeloid-related protein-8 (Mrp8, S100A8), are identified as strong activators of the innate Toll-like receptor 4 signaling pathway, resulting in the expression of pro-inflammatory cytokines
In the current study, exercise did not significantly increase large intestinal permeability. It could be argued that the large intestine is better protected against episodes of hypoperfusion, or that blood flow in the large intestine is better maintained to prevent the translocation of bacteria and their products from the lumen to the circulation. The latter is suggested by a study of Murray et al. that found a decrease in rectal mucosal blood flow of almost 30% after acute physical stress
In this study, we show the development of intestinal damage due to physiological hypoperfusion. Our data imply that the healthy human small intestine can withstand transient hypoperfusion, since the observed enterocyte injury did not produce GI symptoms in our healthy individuals. However, prolonged episodes of hypoperfusion may cause more severe epithelial injury that may give rise to abdominal symptoms, as observed in endurance trained athletes
In conclusion, our study demonstrates that a single, 60-minute bout of moderate-to-high intensity exercise is accompanied by splanchnic hypoperfusion and transient epithelial integrity loss in healthy young males, reflecting the ability of the gut to withstand to episodes of physiological stress with minor consequences.
This study was approved by the medical ethical committee of Maastricht University Medical Centre+, and conducted in accordance with the Declaration of Helsinki (revised version, October 2008, Seoul).
Healthy male volunteers were recruited via posters at the University. All volunteers spent 4 to 10 hours per week performing endurance sports as part of their normal life style. The volunteers had no abdominal complaints during daily activities, had not taken any medication for at least one month prior to participation, had no history of GI disease, and had had no abdominal surgery. Volunteers were informed about the nature and risks of the experiments. Written consent was obtained 5 days before the experiments.
Participants had a mean age of 23.6±0.7 years and body mass index of 21.0±0.4 kg/m2. Subjects' baseline characteristics and maximum workload capacity did not differ between the study groups (
GI perfusion | Intestinal damage | GI permeability | p-value | |
Group size (no. of participants) |
9 | 15 | 6 | |
Age (years) | 24.2±1.0 | 23.8±0.8 | 25.0±0.6 | 0.67 |
Height (m) | 1.79±0.02 | 1.80±0.02 | 1.79±0.00 | 0.79 |
Weight (kg) | 68.4±2.8 | 68.1±1.7 | 66.8±3.8 | 0.92 |
Body Mass Index (kg/m2) | 20.3±0.6 | 21.0±0.5 | 20.8±0.7 | 0.68 |
Total body: fat percentage (%) | 11.6±0.7 | 11.9±0.8 | 12.5±0.6 | 0.81 |
fat mass (kg) | 8.0±0.6 | 8.1±0.6 | 8.4±5.8 | 0.92 |
fat free mass (kg) | 61.1±2.6 | 60.3±1.5 | 59.4±3.8 | 0.91 |
Both legs: fat percentage (%) | 12.6±0.8 | 12.6±0.9 | 13.5±0.6 | 0.77 |
fat mass (kg) | 3.1±0.3 | 3.1±0.3 | 3.3±0.3 | 0.87 |
lean mass (kg) | 20.2±1.2 | 20.5±0.7 | 20.1±1.5 | 0.97 |
Maximum workload (W/kg) | 5.0±0.1 | 5.2±0.1 | 5.3±0.1 | 0.33 |
Data are presented as mean ± SEM.
Of all study participants, three subjects participated in all three substudies, and four subjects participated in two substudies. Hence, in total, 20 individuals took part in the study.
Prior to the experiments, maximal workload capacity was assessed on a stationary cycle ergometer (Lode Excalibur, Groningen, the Netherlands), during electrocardiographic monitoring (MAC 5500, GE Medical Systems, Freiburg, Germany) to exclude cardiologic abnormalities. A dual energy X-ray absorptiometry scan (DXA™, Hologic Inc., the Netherlands) was performed to assess body composition.
Test subjects recorded dietary intake 2 days prior to the first experimental day, and maintained dietary intake as recorded for subsequent test days to prevent dietary influence. Participants were not allowed to consume alcohol or caffeine 2 days prior to each test day. Moreover, participants maintained normal activities of daily living, but refrained from strenuous physical activity. The evening before each test day, participants received a standardized meal (1.7 MJ, consisting of 62.6 g carbohydrate, 18.9 g protein and 7.9 g fat).
Subjects were tested after an overnight fast. The test design is depicted as
Gastric air tonometry was performed in 9 participants. The evening before and 1 hour prior to the experiment, participants ingested 150 mg ranitidine orally (GlaxoSmithKline, Zeist, the Netherlands) to suppress gastric acid production, which interferes with carbon dioxide, thereby affecting tonometry results
I-FABP, a sensitive marker of intestinal cell damage, was used to determine small intestinal damage
Plasma markers for tissue damage and inflammation were analyzed from every subject who completed the 1-hour exercise bout, irrespective of the substudy in which he participated (
Liver injury was assessed by L-FABP, ALT, and AST. Plasma L-FABP is a sensitive indicator of hepatocellular injury
NAG was determined using an enzyme colorimetric assay according to the manufacturer's instructions (HaemoScan, Groningen, the Netherlands). Concentrations were normalized to urinary creatinine values, measured by routine assays at the Clinical Chemistry Laboratory and expressed as Units/mmol creatinine.
GI permeability was determined using a mix of 1 g lactulose (Centrafarm, Etten-Leur, the Netherlands), 1 g sucralose (Brenntag, Sittard, the Netherlands), 1 g erythritol (Danisco, Copenhagen, Denmark), 1 g sucrose (Van Gilse, Dinteloord, the Netherlands), and 0.5 g L-rhamnose (Danisco) dissolved in 150 ml tap water, in a subset of 6 participants.
The study design, depicted in
Combined HPLC (Model PU-1980 pump, Jasco Benelux, Maarsen, the Netherlands) and mass spectrometry (Model LTQ-XL, Thermo Electron, Breda, the Netherlands) were used to determine urinary and plasma sugar concentrations. Lactulose, rhamnose, and L/R ratio were determined in urinary samples to assess small intestinal permeability. Since physiological, transient permeability changes were expected, and our novel permeability assay allowed the detection of sugar probes in plasma, we also determined lactulose and rhamnose concentrations, and L/R ratios in plasma. Both lactulose and rhamnose are degraded in the colon. Therefore, the ratio between two inert sugar probes, sucralose and erythritol, was determined in the 5–24 h urine to assess large intestinal permeability. Since we did not draw blood in this time period, we did not perform large intestinal permeability analysis in plasma.
Circulating myeloperoxidase (MPO) and calprotectin, antimicrobial proteins that are excreted from neutrophils upon activation, are considered early parameters of systemic inflammatory activation
Endotoxin core antibodies in plasma were determined using EndoCab ELISA (Hycult Biotechnology, Uden, the Netherlands; detection window 0.13–8.0 IgG median units (GMU)/mL) according to the manufacturer's instructions in samples taken before and after cycling.
Statistical analysis was performed using GraphPad Prism (Version 5.00, GraphPad Software for Windows, San Diego California, USA). Normality of all data was verified by the Kolmogorov-Smirnov test. All normally distributed data are presented as mean ± standard error of the mean (SEM), not normally distributed data as median and range. Outliers identified using Grubbs analysis were excluded from statistical analysis. Continuous data were analyzed using repeated measures analysis of variance with Tukey's or Dunnett's post-hoc test for multiple comparisons. Basal and post exercise values were compared using the non-parametric Wilcoxon signed rank test, considering the small number of study participants. Within-person correlations between gapg-apCO2 and plasma I-FABP, and plasma I-FABP and plasma L/R ratio were computed by normalizing both data sets, which enables the assessment of the pure association of both variables by calculating the Spearman correlation coefficient rS. Linear regression was used to visualize the correlation. A p<0.05 was considered statistically significant.
(PDF)
(PDF)
We gratefully acknowledge Dr. J. Kolkman, Dr. M. Poeze, Mr. J. Senden, Dr. J. Conchillo, Mr. B. Boonen, and Ms. N. Claes for their advice and technical assistance, and Dr. J. Dekker, Dr. J. Derikx, and Mr. J. Grootjans for critical revision of the manuscript, and all volunteers who contributed to the work reported in this article.