INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

DURING HEAVY PHYSICAL EXERCISE, up to 80% of the increase in cardiac output may be directed to the working muscles to match the increased metabolic demands, thereby curtailing blood flow to splanchnic organs (43). With the use of Doppler ultrasonography, a decrease in splanchnic blood flow of up to 50% has been observed after physical exercise (32, 34). The consequences of such a decrease in blood flow are unclear but may lead to gut mucosal ischemia as observed in animal models of shock (16).

It is well established that strenuous exercise may lead to gastrointestinal abnormalities and cause symptoms such as abdominal pain or discomfort, nausea, vomiting, and diarrhea. Numerous studies have focused on gastrointestinal disorders during exercise (2, 21, 24, 25, 28, 29, 33, 35, 41). The majority of these studies addressed the subject of gastrointestinal motility, but few have investigated the possible role of ischemia as the underlying cause (3, 9, 13, 22, 36, 38, 42). This may be partly explained by the lack of an accurate test to assess actual ischemia independent of metabolic rate and absolute perfusion.

Changes in perfusion per se do not necessarily indicate changes in oxygen supply-to-demand balance, when demand decreases as a consequence of fasting, for instance. The luminal intragastric PCO₂ (Pg_CO2) as measured by tonometry, however, is considered to reflect the mucosal oxygen supply-to-demand balance (18) so that hypoperfusion increases the intragastric-arterial PCO₂ gradient (PCO₂) following a decreased washout and an increased production of CO₂ in ischemic tissue by liberation of CO₂ from HCO buffering of anaerobically produced metabolic acids (11). In fact, heavy exercise may result in an increase of PCO₂ (31). The latter study, however, does not give insight into the level of exercise inducing gastric mucosal ischemia. This is of importance because exercise tonometry has been suggested in a previous paper (17) to be a noninvasive diagnostic tool for stenotic splanchnic vascular disease, but the specificity of abnormal exercise tonometry for vascular disease should be confirmed by normal tonometric values at a similar level of exercise in healthy volunteers with normal splanchnic blood vessels. The cited study was done with the help of the slow manual saline tonometry system; however, the introduction of air tonometry thereafter has allowed for more rapid and accurate Pg_CO2 tonometry (20).

In consideration of the above data, we hypothesised that maximal rather than submaximal exercise in healthy volunteers with normal splanchnic blood vessels causes gastric mucosal ischemia as judged from intragastric air PCO₂ tonometry.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Subjects

Duplex Sonography

Tonometry

A radial artery catheter was inserted in the nondominant arm to allow blood sampling.

Exercise Testing

Protocol

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Schematic time frame of the exercise tonometry study. PgCO2, intragastric PCO2; BL1, baseline intragastric-arterial PCO2 gradient [(PCO2) measurement at time (t) = 10 and 20 min]; EX1, submaximal exercise period; RC1, recovery measurement done at t = 40 min; BL2, baseline PCO2 measurements at t = 70 and 80 min; EX2, maximal exercise period; RC2, recovery measurement at t = 100 min.

Statistics

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Duplex Sonography

Exercise Tonometry

Cardiopulmonary measurements. There were no differences in HR, O₂, and RER between the two baseline episodes (Table 1). At the end of both exercise stages, all cardiopulmonary parameters differed from baseline. HR increased from 78 ± 14 at baseline to 162 ± 6 beats/min (83 ± 3% of HR_max) after EX1 (P < 0.0001) and 189 ± 8 beats/min (97 ± 4% of HR_max) after EX2 (P < 0.0001). The O₂ rose to 91 ± 23% of predicted O_2 max at the end of EX1 and to 130 ± 20% of predicted O_2 max at the end of EX2 (P < 0.0001 for both vs. baseline).

Blood parameters. There were no differences between the two baseline episodes. Pa_CO2 did not change after EX1 but decreased after EX2 and remained below baseline in RC2. The lactate increased and the bicarbonate and base excess levels decreased during both exercise periods but more during EX2 than during EX1.

Tonometry. The baseline Pg_CO2 did not differ between both baseline periods. During EX1, Pg_CO2 increased, whereas the PCO₂ did not change significantly (Fig. 2). During EX2, Pg_CO2 and PCO₂ increased. During RC2, Pg_CO2 returned to baseline but PCO₂ remained elevated over baseline. The mean Pg_CO2 in both baseline periods was 4.9 ± 0.2 kPa, and the coefficient of variation was 4.3%. The PCO₂ during both baseline periods was 0.2 ± 0.4 kPa, with an upper limit of normal (mean + 2SD) of 0.6 kPa.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   PCO2 in the different test episodes. Values are means ± SD. *P < 0.05 vs. baseline value. P < 0.001 for EX1 vs. EX2.

Tonometry variables vs. exercise level. A positive correlation was observed between both Pg_CO2 and PCO₂ with exercise level as indicated by serum lactate at the end of the exercise stage (r = 0.50, P < 0.05 and r = 0.76, P < 0.0001, respectively). No increase in PCO₂ over the normal upper threshold of 0.8 kPa (19) was seen during exercise with lactate levels below 8 mmol/l, whereas in five of six tests, where resulting lactate levels were above 14 mmol/l, the PCO₂ exceeded this threshold. In Fig. 3, the relationship between serum lactate level at the end of the exercise stage and PCO₂ is shown. With the use of bilinear regression analysis, a lactate threshold value of 12 mmol/l was calculated, above which PCO₂ started to rise (r² = 0.73).

View larger version (8K): [in this window] [in a new window]   Fig. 3.   PCO2 vs. lactate level after 10 min of exercise.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results of the present study show that, in subjects with normal splanchnic vasculature, gastric ischemia can develop after only 10 min of maximum physical exercise. The development of gastric ischemia strongly depended on exercise intensity. Furthermore, our study demonstrates that air tonometry can be used for assessment of the adequacy of gastric mucosal perfusion during physical exercise.

The bilinear pattern of the gastric-blood PCO₂ gradient during exercise, with a marked increase above a threshold exercise intensity, is in agreement with studies indicating that tonometric measurement of gastric intramucosal PCO₂ is a reliable index of the adequacy of splanchnic mucosal perfusion (7, 18). If tonometry results depend on blood flow only, a linear pattern between the PCO₂ gradient and lactate should have been observed (14). In contrast, it was previously shown in animal models that a small reduction in gastrointestinal blood flow did not influence Pg_CO2, but, when the blood flow fell to <50% of baseline, hypoperfusion caused an increased Pg_CO2 in parallel with a fall in tissue O₂ tension and consumption (16, 40). Indeed, the gastric-blood PCO₂ gradient may be the most sensitive and specific tonometric parameter of gastrointestinal perfusion, independent of systemic metabolic and respiratory changes (18).

The increase in PCO₂ during maximal exercise is partly caused by a decrease in Pa_CO2. The latter is the result of increased alveolar ventilation in response to metabolic acidosis due to anaerobic muscular metabolism (12). However, in mechanically ventilated patients, it has been shown that changes in Pa_CO2 are closely and swiftly followed by changes in Pg_CO2, thereby not influencing PCO₂ (23). However, even if the decrease in Pa_CO2 is neglected, the relationship between the increase in Pg_CO2 during exercise and arterial lactate level shows a similar bilinear pattern as PCO₂ vs. lactate level [i.e., no increase of Pg_CO2 up to a threshold lactate level of 12 mmol/l and an increase above this lactate threshold (r² = 0.36)].

Our present study confirms the findings of the study of Nielsen et al. (31), which showed that gastric ischemia occurred during heavy exercise in trained rowers at exercise levels close to their maximum aerobic capacity. Using fluid tonometry, they observed, similar to our present study, a gradual increase in Pg_CO2 during exercise with a concomitant decrease in Pa_CO2. A direct comparison of both studies is difficult because the Nielsen study reported on the previously advocated but now largely abandoned intracellular pH (pH_i), a calculated value incorporating arterial bicarbonate levels as well. With profound metabolic acidosis, pH_i might be a poorer predictor of mucosal ischemia than PCO₂. Another potentially confounding factor might have been insufficient acid suppression, since one of their baseline Pg_CO2 already measured up to 8.5 kPa. Similar to previous studies by our group (19), we used high-dose intravenous ranitidine to ensure adequate acid suppression in our subjects.

Automated air tonometry offers several advantages over manual fluid tonometry, including superior accuracy, speed, and reproducibility (18). It is less laborious and involves less error-prone steps, and the faster diffusion of CO₂ in air results in shorter equilibration times. In our study, air tonometry may have slightly underestimated Pg_CO2 values leading to negative PCO₂ baseline values in the majority of measurements (28 of 40 baseline measurements). In one subject in our study, all Pg_CO2 values were lower than Pa_CO2 values resulting in negative gradients, even at maximal exercise. The most likely explanation is the occurrence of air swallowing causing dilution of intragastric gas.

The PCO₂ threshold for anaerobic metabolism is still unclear (18). The range of normal Pg_CO2 and PCO₂ in our study is in close agreement with the study by Creteur et al. (4) on air tonometry showing a normal upper limit of 0.8 kPa for the gradient. Schlichtig and Bowles (40) demonstrated that a fall in tissue O₂ tension and consumption, development of anaerobic metabolism, and production of lactic acid occurred at a PCO₂ of 3.5 kPa or greater, a value that can be regarded as the critical gradient. Thus it may be questioned whether increased PCO₂, as shown in this study, indeed indicates ischemia rather than hypoperfusion of the gastric mucosa. Nevertheless, we cannot exclude ischemia and anaerobic metabolism in the gastric mucosa even at gradients lower than the critical values reported by Schlichtig and Bowles. The 10-min measurement interval of air tonometry may have resulted in an underestimation of the actual peak Pg_CO2 during exercise periods of 10 min. Indeed, the first minutes of the exercise periods were used to gradually increase the workload. Therefore, anaerobic CO₂ production could have occurred only in the last minutes of the exercise episode and not have lasted long enough for full PCO₂ equilibration, resulting in a measured Pg_CO2 that underestimates the actual intraluminal PCO₂ at the end of the exercise. Moreover, exercise-induced hypoperfusion might result in patchy ischemia with anaerobic areas surrounded by still well-perfused areas similar to the ischemic pattern demonstrated in hypovolemic shock (6, 27).

An alternative explanation for increasing PCO₂ after strenuous exercise, other than splanchnic vasoconstriction and ischemia, could be an increase in intra-abdominal pressure, which leads to increased wall tension in the digestive tract, reduced mucosal flow and ischemia. Indeed, in pigs it was demonstrated that a prolonged increase in intra-abdominal pressure of 1.5-3 kPa may decrease mucosal blood flow by 30-40% (5). Similar intra-abdominal pressures have been measured during exercise (15). However, during exercise, the intra-abdominal pressure is not continuously elevated but is changing during each respiratory cycle between ~0 and 3 kPa; it is unlikely that these short (1-2 s) periods of intra-abdominal peak pressure will lead to mucosal ischemia. Moreover, if gastric mucosal ischemia would have been caused by an increase in abdominal pressure during exercise, a rapid normalization in the recovery phase would be expected. In the recovery phase of the present study, however, PCO₂ remained elevated in four subjects and increased even further in one subject. This result is in agreement with the finding that, during recovery after exercise, the splanchnic flow may be impaired for as much as 30 min (34).

An interindividual difference in the response of PCO₂ to the two exercise intensities was noted. In three subjects, the PCO₂ after maximal exercise (with lactate levels increasing to 14.3 mmol/l) was not, or only slightly, higher than after submaximal exercise and still well below the normal threshold of 0.8 kPa. In contrast, in one subject, PCO₂ was already elevated after submaximal exercise (lactate 8.4 mmol/l) and increased greatly after maximal exercise (lactate 15.2 mmol/l) resulting in the highest PCO₂ of all subjects. Although splanchnic arterial abnormalities were excluded by duplex sonography, differences in microvascular anatomy and physiology might very well explain this interindividual susceptibility to gastric mucosal ischemia. This might result from differences in training status as all subjects were recreationally active to a different degree, although none was a competitive athlete (10).

Our results demonstrate a threshold in exercise intensity, as judged from the arterial lactate level, above which a marked increase in PCO₂ was seen as a reflection of gastric mucosal ischemia. This finding of a lactate threshold in individuals with normal splanchnic vasculature, beyond which gastric ischemia may develop, has important implications when exercise gastric tonometry is used as a diagnostic tool in patients suspected of having splanchnic arterial disease. In these patients, in contrast to control patients, exercise of even moderate intensity has been shown to lead to gastric ischemia as judged from a rise in the tonometric Pg_CO2 (17). Our current results in healthy volunteers suggest that, for optimal performance of the test in patients and to prevent false positive results, it is mandatory to continuously monitor and, if necessary, adjust the exercise intensity to keep the lactate level below 8 mmol/l.

In many studies, exercise led to a variety of gastrointestinal abnormalities. The etiology of these abnormalities is still unclear, however. Exercise has been shown to lead to delayed liquid gastric emptying (21, 25, 35), with the magnitude of the delay depending on the exercise intensity (29). With the use of ultrasonography, it has been shown that impaired motility may be explained by exercise-induced closure of the pylorus and a decreased gastric antral area (2). Exercise also has been shown to affect intestinal postprandial motor activity (41). The intestinal epithelial barrier function may be impaired because heavy exercise may result in increased intestinal permeability and impaired water absorption (24). Apart from these functional alterations, heavy exercise, especially marathon running, is also known to cause gastrointestinal blood loss, gastritis, and colitis, which are at least partially ascribed to gastrointestinal ischemia (3, 9, 13, 22, 36, 38, 42). Although often suggested as playing a key role in the development of exercise-induced gastrointestinal abnormalities, gastrointestinal hypoperfusion has not yet been proven to be the cause of these exercise-induced abnormalities. Several investigations (1, 32, 34, 37, 43) that used Doppler ultrasound and thermodilution techniques have shown an exercise-induced decrease of splanchnic blood flow but failed to address the issue of metabolic demand. Our present study using air tonometry shows that, in subjects with normal splanchnic vasculature, gastric ischemia may indeed develop early during maximum physical exercise and that the development of gastric mucosal ischemia is strongly dependent on the exercise intensity.