Both gastric emptying and intestinal absorption are important for maintaining fluid homeostasis and for supplementing endogenous carbohydrate stores during exercise. Depending upon environmental conditions and the metabolic rate sustained during exercise: (a) sweat rates can exceed 1.0 l·h-1, resulting in body weight losses exceeding 2-4%; (b) blood glucose oxidation can exceed 1.0 g·min-1(2); and (c) temperatures can exceed 39°C(1,7,17). Thus, during intense exercise relatively high rates of fluid and carbohydrate replacement may be essential to prevent excessive dehydration, glycogen depletion, hyperthermia, and decrements in performance. To attain high rates of fluid and carbohydrate replacement, ingested nutrients must first be rapidly emptied from the stomach and then rapidly absorbed by the intestine.
Previous investigations have demonstrated that repeated ingestion of small volumes of water or dilute (up to 8%) carbohydrate-electrolyte solutions (CES) can maintain relatively large gastric volumes, enhance gastric function, and thereby provide relatively high delivery rates of both fluids (15-20 ml·min-1) and carbohydrates (30-60 g·h-1) to the small intestine (16,17,19,22). In general, gastric emptying of water and dilute CES increase with increasing gastric volume; however, very high gastric volumes may inhibit gastric emptying and elicit sensations of gastrointestinal distress(5,16). Given that gastric function is largely controlled by the rate of nutrient delivery to the duodenum(12) and that gastrointestinal distress can be caused by gastric fullness or intestinal malabsorption, the present investigation sought to determine if the rate of intestinal absorption could be a limiting factor in oral rehydration. In contrast to knowledge of gastric function, little is known about the maximum absorptive capacity of the small intestine(25).
The present study was designed to test the upper limit of intestinal absorption of a 6% glucose-electrolyte solution (GES) in healthy males at rest. In Study I gastric emptying (GER) of the test solution was determined after an initial load and while sustaining the load by repeated infusion of large solution volumes into the stomach. This protocol was designed to elicit near maximal gastric emptying. In Study II a test solution formulated to match duodenal composition of the Study I solution was perfused into the proximal small intestine at rates equal to its predetermined GER and at two higher rates equal to GER increased by 7.5 or 15.0 ml·min-1, respectively. By increasing the perfusion rate beyond the measured GER, we sought to determine the effect of increasing water and glucose load on duodenojejunal absorption and the maximal capacity of the proximal small intestine to absorb the test solution.
Subjects. Six healthy male volunteers provided informed consent and consent and served as subjects. Physical characteristics (mean ± SE) were age 25.4 ± 1.4 yr, weight 82.0 ± 2.9 kg, and height 182.3 ± 2.9 cm. Each subject completed a medical examination and reported no history of gastrointestinal abnormalities or disease. All experimental procedures were approved by the University of Iowa Human Use Committee.
Experimental design. This investigation consisted of two consecutive studies. All experiments were performed at the same time of day after an overnight fast, and experiments in the same individual were separated by at least 7 d. In Study I, the gastric emptying rate and duodenal composition of the test solution were determined during the repeated infusion(via a nasogastric tube) of large volumes into the stomach for 50 min. In Study II, the test solution of Study I was modified to match duodenal composition determined in Study I. This solution was perfused via a triple lumen tube into the proximal small intestine (10-cm mixing segment and 40-cm test segment) at three different perfusion rates for 50 min. Intestinal water and solute absorption were determined at perfusion rates equal to the predetermined GER (Study I), and at two higher rates equivalent to the GER increased by 7.5 and 15.0 ml·min-1, respectively.
Study I. Following a 10-12 h overnight fast, subjects were intubated with both a nasogastric tube (French Levine,12 gauge) and a triple-lumen tube (Arndorfer, Greendale, WI) at the Degestive Disease Center of the University of Iowa Hospitals and Clinics. Under fluoroscopic guidance, the nasogastric tube was placed into the gastric antrum while a smapling port of the multilumen tube was positioned within the proximal small intestine(≈5-10 cm distal to the pylorus). Intubations generally required ≈60-90 min with fluoroscopy time 5-15 s.
After returning to the Exercise Physiology Laboratory, subjects were seated for 30 min, and the stomach was washed with repeated(5-10) infusions of 200 ml of distilled H2O until the recovery solution was clear. Immediately following the final wash, a single large bolus (9.1 ml·kg-1 BW;≈750 ml) of chilled (10-15°C) test solution was passed into the stomach by gravity flow. The test solution contained 6% glucose (339 mmol·l-1) with 20 mEq·l-1 Na+ and 3 mEq·l-1 K+, osmolality 401 mOsm·kg-1. Additional small serial infusions (2.2 ml·kg-1 body weight; ≈ 180 ml) were given by the same route every 10 min thereafter. Gastric volume, residual volume, emptying, and secretion, as well as composition of the solution, emptied into the duodenum were determined at 10-min intervals for 50 min.
Gastric emptying and secretion were determined using the double sampling technique of Geroge (6). Briefly, the concentration of phenol red dye was determined in gastric samples obtained before and after consecutive infusions of test solutions containing 0, 12.5, 50, 100, and 200 mg·l-1 phenol red. Prior (1 min) to each 10-min interval, the stomach contents were mixed, and a 5 ml gastric sample was aspirated. Additional test solution (2.2 ml·kg-1 body weight) was then infused, stomach contents were mixed again, and a second 5-ml gastric sample was collected. Mixing was performed with a 60 cc syringe, required ≈ 1 min to complete, and consisted of repeated (11 times) withdrawal and instillation of 20-30 ml of gastric contents. To determine the effect of gastric and duodenal secretions and absorption on the infused test solution, intestinal samples were obtained from the multi-lumen tube at 10-min intervals. Samples were stored at -20°C until analysis.
Study II. Each subject completed a separate experiment in which the modified test solution was perfused into the duodenojejunum at three different perfusions rates. Following an overnight fast, the same subjects(N = 6) were intubated with a triple-lumen tube as described in Study I. The proximal sampling site was positioned ≈15-20 cm distal to the pylorus sphincter, and the distal sampling site was 40 cm from the proximal sampling site and extended about 25 cm into the proximal jejunum. Following intubation, a needle catheter fitted with a heparin lock was inserted into a superficial forearm vein.
Subjects rested quietly in a seated position for 15 min. The“modified” test solution was then perfused for 50 min at a rate equal to predetermined GER (measured in Study 1). A 20-min equilibration perfusion was followed by a 30-min sampling period. During this 30 min, intestinal fluid was collected from the proximal sampling site at a rate of 1 ml·min-1 and from the distal site by constant syphonage. Following the initial 50-min perfusion, subjects were allowed to walk and stretch for 20 min. Perfusion at the next higher perfusion rate (7.5 ml·min-1) above the GER was then begun, and the same protocol was followed. The final protocol was to perfuse at the GER increased by 15.0 ml·min-1.
Net intestinal absorption of water and solutes were determined using the techniques and calculations described by Cooper et al. (3) and Gisolfi et al. (9). Calculations of net water and solute flux were conducted using samples collected after the 20-min equilibration period. All test solutions contained 1.0 mg·ml-1 polyethylene glycol 4000 (PEG) as a nonabsorbable marker.
Analyses. Phenol red in gastric samples was determined by spectrophotometric analysis according to Schedl (24). Gastrointestinal samples were analyzed for glucose (glucose analyzer model 23A, Yellow Springs Instrument, Yellow Springs, OH), osmolality (Model 5300, Wescor, Logan, VT), and Na+ and K+ (flame photometer model 143, Instrumentation Laboratory, Lexington, MA). Analysis of PEG in intestinal samples was performed by the turbidometric assay as modified by Malawer and Powell (13).
A one-factor ANOVA with repeated measures was employed in Study I to determine: 1) the effect of time on gastric volume, gastric emptying rate, residual volume, and gastric secretion, and 2) changes to the beverage composition in the stomach and duodenum. Significant differences (P< 0.05) were identified using the Scheffe post-hoc test. In Study II, a two-factor ANOVA with repeated measures was used to determine the effect of infusion rate and time in a given experiment on water, glucose, sodium, and potassium flux at each infusion rate. Statistical significance was set atP < 0.05. All values reported are mean ± SE.
Study I. Loading the washed stomach with a 750 ml of the 6% GES test solution followed by serial infusions of ≈ 180 ml at 10 min intervals(total volume infused ≈ 1.47 l) resulted in a high rate of gastric emptying: 19.5 ± 1.4 ml·min-1 (mean ± SE; range for the six subjects 18.8-21.0 ml·min-1). This loading procedure was tolerated well; no subject complained of GI distress. The relatively high gastric and residual volumes maintained a high gastric emptying rate throughout the experiment. There were no statistically significant differences in gastric secretion, but gastric and residual volumes did fall from their initial high values (Table 1). This decline no doubt represents the stomach adjusting to the initial high bolus volume; however, despite these differences during the 50-min test period gastric volumes averaged 713 ± 58 ml and final gastric residual volume was 507 ± 26 ml following the repeated infusions of the ≈ 180 ml of the 6% GES. The stomach delivered 0.98 1·50 min-1 of fluid and 1.2 g·min-1 (4.8 Kcal·min-1) of glucose to the duodenum.
Composition of the test solution was altered by the stomach and duodenum(Fig. 1). Compared with the original solution, [K+] increased and glucose concentration and osmolality decreased in the stomach. Na+ concentration of the test solution increased in the duodenum and glucose concentration and osmolality were further reduced(Fig. 1).
Study II. The test solution used in Study II(Table 2) matched duodenal composition determined in Study I and was perfused into the proximal small intestine at the flow rate of the subject with the greatest GER (control: 21 ml·min-1), GER increased by 7.5 ml·min-1, and GER increased by 15.0 ml·min-1. Both duodenal composition and GER were predetermined in Study I.
The test solution was well tolerated when perfused into the proximal small intestine at all perfusion rates. Two subjects complained of mild abdominal cramping at the end of perfusion at the highest (36 ml·min-1) rate. Water flux values during perfusion of the test solution were similar for control (21.0 ml·min-1) and at 28.5 and 36.0 ml·min-1: - 12.9 ± 2.3, - 11.3 ± 3.0, and - 12.7± 6.4, respectively (Fig. 2).
Intestinal glucose absorption during perfusion with the modified test solution was similar at all perfusion rates (Fig. 3). Likewise, net intestinal absorption of Na+ and K+ were independent of perfusion rate.
This investigation showed that an initial large load of a GES followed by repeated infusion of smaller volumes of GES into the stomach not only maintained relatively high gastric volumes across experimental time but also elicited a high GER. Gastric volume was maintained for 50 min by providing the test solution initially as a single large bolus (≈750 ml) followed by additional small serial feedings (≈180 ml) every 10 min.
Although this study was performed at rest, the data also apply to exercise conditions based on the findings in the literature. Costill and Saltin(4) reported that a single 600- or 800-ml feeding of a 2.5% glucose solution elicited a GER of ≈25 ml·min-1 during 15 min of rest. Ryan et al. (22) observed GER exceeding 17 ml·min-1 in cyclists (3 h at 60% peak ˙VO2) who consumed 5% carbohydrate solutions at a rate of 350 ml every 20 min. Similarly, Mitchell and Voss (16) reported GER of ≈ 12, 16, and 19 ml·min-1 in cyclists (2 h at 70%˙VO2max) who consumed a 7.5% carbohydrate solution at rates of≈200, 300, and 400 ml every 15 min. Finally, Rehrer et al.(20,21), using runners and cyclists, reported high GER for water and dilute CES when these solutions were consumed as an initial large bolus followed by repeated small feedings. Together with the present investigation, these findings support the proposal (18) that gastric emptying is strongly influenced by the total volume ingested, the average gastric volume maintained, and the drinking pattern used. These data also suggest that, for water and dilute (up to 8%) CES, maximum GER may be achieved by adopting a drinking pattern that sustains gastric volumes at values of ≈400 to 600 ml, i.e., at gastric volumes that are presumably close to values that produce gastric fullness and/or discomfort(5,16).
The relatively high GER achieved in this study provided a high delivery rate of both fluid and carbohydrate to the small intestine. Gastric emptying of the test solution translated to a fluid delivery rate of 1.17 l·h-1 and a caloric delivery rate of 4.8 Kcal·min-1, respectively. Assuming complete absorption during passage along the gastrointestinal tract, these gastric emptying rates of fluids and calories appear to be sufficient to meet the fluid and carbohydrate needs of most endurance athletes (5). Previous studies have shown that gastric fluid and caloric delivery rates are influenced by both the volume and energy density of the test solution(10). For example, Mitchell et al.(15) reported gastric fluid and caloric delivery rates, respectively, of ≈605, 590, 525, and 445 ml·h-1 and 0, 2.4, 4.2, and 5.3 Kcal·min-1 in cyclists (2 h at 70%˙VO2max) who consumed water, and a 6%, 12%, and 18% carbohydrate solution at a constant rate of 150 ml every 15 min. Mitchell and Voss(15) also estimated gastric fluid rates of 720, 960, and 1170 ml·h-1 and caloric delivery rates of 4.2, 6.4, and 8.5 Kcal·min-1 in cyclists who consumed a 7.5% carbohydrate solution at varying rates of ≈200, 300, and 400 ml every 15 min. Thus, the high gastric volumes attained by repeated stomach infusions explain the high gastric delivery rates of both fluids and calories observed in the present study.
Gastric secretions did not alter the electrolyte concentrations of the test solution. However, they may have contributed to the reductions in glucose concentration and osmolality. In contrast, the duodenum increased[Na+], but had little effect on[K+]. This may be explained by Na+ moving down a large electrochemical gradient (plasma Na+ at 140 mEq·l-1 compared with ≈15 mEq·l-1 in chyme), whereas the gradient for K+ is small (plasma K+ at 4.5 mEq·l-1 compared with ≈4 mEq·l-1 in chyme). We attribute the rise in [Na+] within the duodenum to Na+ secretion rather than to intestinal absorption of water because K+ did not rise significantly. The shift in osmolality toward isotonicity in the duodenum agrees with Miller et al. (14) and Shi et al.(26), who reported that luminal contents of the distal duodenum has an osmolality of 300 mOsm·Kg-1. These findings indicate that the composition of a dilute GES is modified little by the stomach but more dramatically by the proximal duodenum.
Gastric emptying is controlled by numerous feedback loops whereby the quantity and quality of digested nutrients is closely monitored such that gastric emptying can be rapidly adjusted to ensure adequate intestinal digestion and absorption (12). In the present study these feedback control loops were bypassed by directly perfusing the test solution into the duodenum at rates equal to the GER of test solutions and at two higher rates. By increasing the perfusion rate beyond the determined GER, we sought to determine the effect of perfusion rate on duodenojejunal absorption, as well as to estimate the maximal capacity of the small intestine to absorb the test solutions.
Maximum fluid absorptive capacity of the small intestine is unknown, but estimates can be derived from absorption rates measured within the duodenojejunum. The duodenojejunum absorbs the major (≈75%) portion of the volume load presented to the small intestine. Assuming a 100-cm length for the duodenojejunum and an absorption rate of 13 ml·h-1·cm-1(8,9) along the entire length, fluid absorption within this segment is estimated to be≈1.3 l·h-1. This estimate of intestinal absorption was obtained from perfusion of the test solution at a rate of 15 ml·min-1, a rate reasonably matched with the GER of this solution. Importantly, this estimate also does not include fluid absorption that occurs within the distal intestine, i.e., distal to the intestinal test segment.
In the present study, perfusion of a 6% GES at its measured GER (21 ml·min-1), and at two higher rates (28.5 and 35.0 ml·min-1) produced similar segmental absorption rates of≈12.3 ml·h-1·cm-1, demonstrating that fluid absorption within the duodenojejunum (≈1.23 l·h-1) may not be enhanced by increasing perfusion rate and that fluid absorption of this solution appears to be closely matched to its determined GER. However, intestinal absorption rates may vary depending upon solution composition and the intestinal segment under study (26). Thus, the similar absorption rates for 6% GES may indicate that when this solution was perfused at the two higher rates greater absorption of this solution may have occurred in the distal intestine (distal to the test segment).
Shi et al. (26) observed duodenojejunum water absorption rates equivalent to 1.17, 1.31, and 1.61 l·h-1 during perfusion of a hypotonic, an isotonic, and a hypertonic 6% CES, respectively, at a perfusion rate of 15 ml·min-1. These investigators noted that intestinal water absorption was greater in solutions that contained more than one transportable sugar. Lambert et al. (11) recently reported duodenojejunal water absorption rates of ≈1.95 or 1.64 l·h-1, respectively, during repeated oral ingestion or during intestinal perfusion of a 6% CES at rates equivalent to ≈20 ml·min-1. Likewise, Ryan et al. (23) reported that repeated ingestion of either water, a 6%, 8%, or 9% CES produced a similar GER of ≈18 ml·min-1, but resulted in markedly different rates of water absorption from the duodenojejunum, the absorption rates were equivalent to ≈1.83, 1.65, 0.69, and 0.18 l·h-1, respectively.
In the present experiment, intestinal glucose absorption during perfusion of the test segment was not altered by perfusion rate. Mean values ranged from 4.3 to 5.6 mmol·h-1·cm-1 for 6% GES perfused at 21 and 28.5 ml·min-1. Assuming a 100-cm length of duodenojejunum and an equal absorption rate along the length, glucose absorption within this intestinal segment was estimated to range from 1.3 to 1.7 g glucose·min-1. This glucose absorption rate is similar to the gastric glucose emptying rate of 1.2 g·min-1. These data show that GER is controlled in a manner that closely matches the rate of gastric glucose delivery with rates of intestinal glucose absorption. This glucose absorption rate appears to be sufficient to meet the needs of most competitive endurance athletes (5).
For competitive athletes requiring high rates of fluid and carbohydrate replacement, it is important that both fluids and carbohydrates can be rapidly emptied from the stomach and then rapidly absorbed from the intestines. The present study provides data regarding near maximal rates of gastric emptying and intestinal absorption. Gastric emptying values obtained during repeated stomach infusion of large volumes of a 6% GES were estimated to provide fluid delivery at a rate of 1.17 l·h-1 to the small intestine and a caloric delivery rate of 4.8 Kcal·min-1, respectively. When perfused into the duodenum at its GER, the test solution produced a fluid absorption of 1.29 l·h-1. The corresponding intestinal glucose absorption rate for the test solution was estimated at 1.3 g glucose·min-1 and was closely matched to its respective gastric glucose delivery rate of 1.2 g glucose·min-1. These data show that the gastrointestinal tract of healthy males displays a high capacity to deliver both fluid and calories at rates that appear to be sufficient to match the fluid and carbohydrate needs of most competitive athletes.
1. Barr, S. I., D. L. Costill, and W. J. Fink. Fluid replacement during prolonged exercise: effects of water, saline, or no fluid.Med. Sci. Sports Exerc.
2. Coggan, A. R. and E. F. Coyle. Reversal of fatigue during prolonged exercise by carbohydrate infusion or ingestion. J. Appl. Physiol.
3. Cooper, H., R. Levitan, J. S. Fordtran, and R. J. Ingelfinger. A method for studying absorption of water and solute from the human small intestine. Gastroenterology
4. Costill, D. L. and B. Saltin. Factors limiting gastric emptying during rest and exercise. J. Appl. Physiol.
5. Coyle, E. F. and S. J. Montain. Carbohydrate and fluid ingestion during exercise: are there trade-offs? Med. Sci. Sports Exerc.
6. George, J. D. New clinical method for measuring the rate of gastric emptying: the double sampling test meal. Gut
7. Gisolfi, C. V. and J. R. Copping. Thermal effects of prolonged treadmill exercise in the heat. Med. Sci. Sports
8. Gisolfi, C. V., K. J. Spranger, R. W. Summers, H. P. Schedl, and T. L. Bleiler. Effects of cycle exercise on intestinal absorption in humans. J. Appl. Physiol.
9. Gisolfi, C. V., R. W. Summers, H. P. Schedl, T. L. Bleiler, and R. A. Oppliger. Human intestinal water absorption: Direct vs indirect measurements. Am. J. Physiol.
10. Hunt, J. N., J. L. Smith, and C. L. Jiang. Effect of meal volume and energy density on the gastric emptying rate of carbohydrates.Gastroenterology
11. Lambert, G. P., R. T. Chang, D. Joensen, et al. Simultaneous determination of gastric emptying and intestinal absorption during cycle exercise in humans. Int. J. Sports Med.
12. Lin, H. C., J. E. Doty, T. J. Reedy, and J. H. Meyer. Inhibition of gastric emptying by glucose depends on length of intestine exposed to nutrient. Am. J. Physiol.
13. Malawer, S. J. and D. W. Powell. An improved turbidimetric analysis of polyethylene glycol utilizing an emulsifier.Gastroenterology
14. Miller, L. J., J.-R. Malagelada, and V. L. W. Go. Postprandial duodenal function in man. Gut
15. Mitchell, J. B., D. L. Costill, J. A. Houmard, W. J. Fink, R. A. Robergs, and J. A. Davis. Gastric emptying: influence of prolonged exercise and carbohydrate concentration. Med. Sci. Sports Exerc.
16. Mitchell, J. B. and K. W. Voss. The influence of volume on gastric emptying and fluid balance during prolonged exercise. Med. Sci. Sports Exerc.
17. Noakes, T. D., K. H. Myburgh, J. Du Plessia, et al. Metabolic rate, not percent dehydration, predicts rectal temperature in marathon runners. Med. Sci. Sports Exerc.
18. Noakes, T. D., N. J. Rehrer, and R. J. Maughan. The importance of volume in regulating gastric emptying. Med. Sci. Sports Exerc.
19. Owen, M. D., K. C. Kregel, P. T. Wall, and C. V. Gisolfi. Effects of ingesting carbohydrate beverages during exercise in the heat. Med. Sci. Sports Exerc.
20. Rehrer, N. J., E. J. Beckers, F. Brouns, F. Tenhoor, and W. H. M. Saris. Effects of dehydration on gastric emptying and gastrointestinal distress while running. Med. Sci. Sports Exerc.
21. Rehrer, N. J., F. Brouns, E. J. Beckers, F. Tenhoor, and W. H. M. Saris. Gastric emptying with repeated drinking during running and bicycling. Int. J. Sports Med.
22. Ryan, A. J., T. L. Bleiler, J. E. Carter, and C. V. Gisolfi. Gastric emptying during prolonged cycling exercise in the heat.Med. Sci. Sports Exerc.
23. Ryan, A. J., G. P. Lambert, X. Shi, R. T. Chang, R. W. Summers, and C. V. Gisolfi. Effect of hypohydration on gastric emptying and intestinal absorption during exercise. FASEB J.
24. Schedl, H. P., D. Miller, and D. White. Use of polyethylene glycol and phenol red as unabsorbed indicators for intestinal absorption studies in man. Gut
25. Schiller, L. R., C. A. Santaana, and J. S. Fordtran. Glucose stimulated Na absorption: relative contribution of glucose-sodium cotransport and solvent drag during total intestinal perfusion.Gastroenterology
26. Shi, X., R. W. Summers, H. P. Schedl, R. T. Chang, G. P. Lambert, and C. V. Gisolfi. Effects of solution osmolality on absorption of select fluid replacement solutions in human duodenojejunum. J. Appl. Physiol.