Journal of Applied Physiology
Vol. 83, No. 1, pp. 204-212, July 1997
SYSTEMIC CIRCULATION AND FLUID BALANCE

Absorption from different intestinal segments during exercise

G. P. Lambert, R. T. Chang, T. Xia, R. W. Summers, and C. V. Gisolfi

Departments of Exercise Science and Internal Medicine, University of Iowa, Iowa City, Iowa 52242-1111

ABSTRACT

INTRODUCTION

MATERIALS AND METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Lambert, G. P., R. T. Chang, T. Xia, R. W. Summers, and C. V. Gisolfi. Absorption from different intestinal segments during exercise. J. Appl. Physiol. 83(1): 204-212, 1997.This study evaluated intestinal absorption from the first 75 cm of the proximal small intestine during 85 min of cycle exercise [63.6 ± 0.7% peak O₂ consumption (O_{2 peak})] while subjects ingested either an isotonic carbohydrate-electrolyte beverage (CHO-E) or a water placebo (WP). The CHO-E beverage contained 117 mM (4%) sucrose, 111 mM (2%) glucose, 18 meq Na⁺, and 3 meq K⁺. The two experiments were performed a week apart by seven subjects (6 men and 1 woman; mean O_{2 peak} = 53.5 ± 6.5 ml · kg¹ · min¹). Nasogastric and multilumen tubes were fluoroscopically positioned in the gastric antrum and duodenojejunum, respectively. Subjects ingested 23 ml/kg body weight of the test solution, 20% (383 ± 11 ml) of this volume 5 min before exercise and 10% (191 ± 5 ml) every 10 min thereafter. By using the rate of gastric emptying (18.1 ± 1.1 vs. 19.2 ± 0.7 ml/min for WP and CHO-E, respectively) as the rate of intestinal perfusion, intestinal absorption was determined by segmental perfusion from the duodenum (0-25 cm) and jejunum (25-50 and 50-75 cm). Water flux was different (P < 0.05) between solutions in the 0- to 25- and 25- to 50-cm segments for WP vs. CHO-E (30.7 ± 2.7 vs. 15.0 ± 2.9 and 3.8 ± 1.1 vs. 11.9 ± 3.3 ml · cm¹ · h¹, respectively). Furthermore, water flux differed (P < 0.05) for WP in a comparison of the 0- to 25- to the 25- to 50-cm segment. Total solute flux (TSF) was not significantly different among segments for a given solution or between solutions for a given segment. There was no difference between trials for percent change in plasma volume. These results indicate that 1) fluid absorption in the proximal small intestine depends on the segment studied and 2) solution composition can significantly effect water absorption rate in different intestinal segments.

duodenum; jejunum; fluid balance; fluid absorption; solute absorption

INTRODUCTION

AN IMPORTANT ASPECT of fluid homeostasis during prolonged exercise is the ability to absorb ingested fluids (19), which occurs primarily in the proximal small intestine (duodenojejunum). Segmental perfusion is the most accepted technique for study of intestinal absorption of different solutions (47); however, the results only apply to the segment studied. Previous investigations (1, 17, 21-23, 26, 29, 39, 41, 46, 54) have differed in the site of perfusion (e.g., duodenojejunum, jejunum, ileum), which has led to difficulty in comparing results and drawing conclusions regarding the efficacy of various solutions. Furthermore, caution must be exercised in applying the results of direct perfusion studies to what would occur if the same solution were orally ingested.

Recently, we developed a technique to simultaneously determine both gastric emptying (GE) and intestinal absorption (28). This technique allows determination of intestinal water and solute flux, while also accounting for individual GE rates and alterations made by the stomach to the solution after its oral ingestion. This technique was employed in the present investigation to determine whether different segments of the proximal small intestine absorb water and nutrients at different rates.

In humans, the duodenum is defined as the portion of intestine from the pyloric sphincter to the ligament of Treitz and is 20-30 cm in length (10, 33). It is highly permeable and presumably the site of maximal fluid movement between the alimentary canal and the blood. It is often cited as the intestinal segment responsible for bringing chyme from the stomach to isotonicity (51). Most segmental perfusion studies bypass this segment and typically evaluate jejunal function beyond the ligament of Treitz. If the purpose of performing segmental perfusion in the jejunum is to evaluate the efficacy of ingesting the same solution, such an extrapolation may be inaccurate. In contrast to the duodenum, the jejunum consists of a less leaky mucosal epithelium with a significantly greater ability to absorb glucose (24). Results of a pilot study from this laboratory (Fig. 1) indicate significantly greater fluid absorption with ingestion of a water placebo (WP) compared with an isotonic 6% carbohydrate-electrolyte (CHO-E) beverage in the duodenum. This contrasts with previous findings (22, 23) in which primarily the jejunum was studied by the segmental perfusion technique. From these data, we questioned whether segmental differences exist for fluid and solute absorption.

Given the large osmotic gradient between water and blood, we hypothesized that absorption of water would be greater in the duodenum (first 25 cm) compared with an isotonic solution. In the jejunum (second and third 25-cm segments), where membrane resistance is greater than in the duodenum but where glucose transport is enhanced (24), we hypothesized an isotonic CHO-E beverage would have an absorptive advantage over deionized water because of CHO absorption and subsequent "solution drag." To test these hypotheses, we studied water and solute absorption of two different solutions (a deionized WP and an isotonic CHO-E solution) in three consecutive segments of the proximal small intestine (~0-25, 25-50, and 50-75 cm distal to the pyloric sphincter).

MATERIALS AND METHODS

Six healthy men and one woman [age 26.0 ± 9.0 yr; peak oxygen consumption (O_{2 peak}) = 53.5 ± 6.5 ml · kg¹ · min¹] participated in this study, which conformed to all the rules and regulations of the University of Iowa Human Use Committee. Each subject gave signed informed consent and received a physical examination before participation. O_{2 peak} was determined by using a graded protocol on an electronically braked cycle ergometer (The Bike, Cybex, Ronkonkoma, NY). Expired gases and ventilation were analyzed with a Q-Plex I metabolic system (Quinton Instruments, Seattle, WA). A workload corresponding to 60-65% O_{2 peak} was subsequently determined and served as the exercise intensity for the experimental trials.

Subjects followed a controlled diet for 16-18 h before each experiment and were asked to limit their physical activity during this time to control preexperiment glycogen levels. The morning of an experiment, the subject arrived at the University of Iowa Hospitals and Clinics at ~5 AM after an overnight fast (at least 8 h) and was orally intubated with a nasogastric (NG) tube (50 in., 14 Fr, Levin) into the gastric antrum and a multilumen tube (construction described below) into the duodenojejunum. During tube placement the subject ingested water (~400-500 ml) to facilitate movement of the tubes and to control for preexperiment hydration status. The NG tube was attached to the multilumen tube with a small piece of thin rubber tubing so that its distal tip was 9 cm from the proximal sampling site of the multilumen tube. The rubber tubing was located 7 cm from the distal tip of the NG tube to allow the NG tube to separate from the multilumen tube in the stomach without passing into the duodenum. Tube placement was verified fluoroscopically, and contrast medium [Hypaque sodium 50%; a brand of diatrizoate sodium injection (USP)] was injected through the proximal sampling port during placement of the tube to ensure that the tube was positioned correctly. When both tubes were in position, the most proximal port of the multilumen tube was located ~5 cm from the pyloric sphincter, and the NG tube was positioned in the distal antrum.

During positioning of the tubes, an 18-gauge, 13/4-in. catheter was placed in a superficial arm vein for blood sampling. After tube placement, the subject walked to the exercise physiology laboratory and sat for 20 min to allow plasma volume to equilibrate. After 20 min, blood and urine samples were collected, a rectal temperature (clinical thermometer), resting heart rate (heart rate monitor; Polar Vantage XL, Polar USA, Stamford, CT), and nude body weight were obtained, and the subject changed into cycling clothes. The subject then immediately mounted the stationary bike, and stomach contents were aspirated through the NG tube.

After stomach aspiration, the subject drank an initial bolus of test solution equaling 20% of the total volume ingested (23 ml/kg body wt). The total volume averaged 1,914 ± 191 ml, and the mean initial bolus was 383 ± 37 ml. Five minutes after ingestion, the subject began cycling for 85 min. Each 10-min interval thereafter, an additional amount of test solution was ingested and equaled 10% of the total experimental volume (191 ± 19 ml). The experimental solutions consisted of either a 117 mM (4%) sucrose, 111 mM (2%) glucose, 17.8 meq/l Na⁺, and 3.1 meq/l K⁺ beverage with an osmolality of 282 mosmol/kgH₂O (i.e., CHO-E), or a deionized WP (osmolality = 1.1 mosmol) flavored to match the CHO-E solution. Each solution also contained 1 mg/ml polyethylene glycol 3350 (PEG). The temperature of the solution, which was given to the subject in a clear graduated flask, was 10-15°C. Experiments were performed in a 22°C environment with a slight breeze (~2 ft /s) produced by a fan placed in front of the subject. Blood samples were drawn every 15 min to determine changes in plasma volume, osmolality, Na⁺, K⁺, and glucose. Heart rate was obtained every 15 min. Rectal temperature, nude body weight, and urine volume were recorded postexercise. Sweat rate was calculated from nude body weight change pre- to postexperiment corrected for fluid ingestion, phenol red (PR) injection, and stomach, intestinal, blood, and urine samples.

Determination of Stomach Volume, GE Rate, and Gastric Secretion

Gastric secretion is calculated as where GS₁₀ is the volume of gastric secretion in the stomach contents after each 10-min period from the time a subject drinks to the time stomach volume is determined 10 min later. Stomach volume is that determined for the 10-min period under study, PR₁₀ is the PR concentration in the stomach before the stomach volume is determined (after 10 min have elapsed since drinking), and PR₀ is the PR concentration in the stomach immediately after the initial drink (10 min before). GE rate is derived by subtracting a given stomach volume from the volume that is in the stomach after the previous ingestion of beverage 10 min before.

Determination of Intestinal Absorption by Segmental Perfusion

Analytical Procedures

Statistical Analysis

The pilot study referred to at the beginning of this study was carried out identically to the methods described above, except that there was only a 0- to 25-cm test segment in the multilumen tube (for specifics, see Ref. 27).

RESULTS

All subjects began the experiments in a euhydrated state on the basis of plasma osmolality values (290 ± 1 mosmol/kgH₂O for both WP and CHO-E). There were no differences in cardiovascular, fluid, or thermoregulatory measurements (sweat rate, percent body weight loss, urine production, rectal temperature, final heart rate) between the two experimental conditions (Table 1). Fluid ingestion offset fluid losses due to sweating and sample collection (see %body weight loss, Table 1).

Table  1.   Comparisons of sweat rate, percent body weight loss, urine production, final rectal temperature, and final heart rate between the 2 experimental conditions Solution Sweat Rate, l/h Body Wt Loss, %  Urine Produced, ml Final Rectal Temperature, °C Final Heart Rate, beats/min WP 1.13 ± 0.10  0.18 ± 0.15  255 ± 97  38.4 ± 0.10  146 ± 7  CHO-E 1.11 ± 0.02  0.18 ± 0.09  166 ± 35  38.5 ± 0.08  146 ± 5 Values are means ± SE; n = 7 subjects. WP, water placebo; CHO-E, carbohydrate-electrolyte beverage.

Mean GE did not differ between the two drinks (18.1 ± 1.1 vs. 19.2 ± 0.7 ml/min for WP and CHO-E, respectively). Mean gastric secretion rates were also not different for the two solutions (26.0 ± 7.1 vs. 23.7 ± 10.8 ml/10 min for WP and CHO-E, respectively). GE was maintained at a steady rate by producing a relatively constant stomach volume. There were no differences over time for stomach volume or GE in either experiment after the 35-min equilibration period (mean stomach volume postequilibration: WP = 264 ± 60 ml; CHO-E = 240 ± 43 ml) (Fig. 2). In terms of intestinal steady-state conditions (after the 35-min equilibration period), no differences were observed in PEG concentrations at a given sampling site over time for a given solution. However, significant differences were observed in mean fluid absorption in the different segments of the small intestine for a given solution (Fig. 3). Moreover, water flux in specific areas of the small intestine was dependent on solution composition (Fig. 3). When the whole 75-cm segment was examined, no difference occurred in water absorption (12.4 ± 1.1 vs. 10.4 ± 1.1 ml · cm¹ · h¹ for WP and CHO-E, respectively) (Fig. 3). Total water absorption for the whole 75-cm segment accounted for 83 and 72% (not sigificant) of the fluid available for absorption (after correction for fluid sampled from the stomach and intestine) for WP and CHO-E, respectively. Of the fluid absorbed, 80 and 82% were retained for WP and CHO-E, respectively, when urine production (Table 1) was taken into account. Total solute flux (TSF) was significantly different (0.8 ± 0.2 vs. 4.2 ± 0.3 mmol · cm¹ · h¹ for WP and CHO-E, respectively; P < 0.05) over the entire 75 cm. TSF did not differ among segments for a given solution, or between beverages within segments, although the CHO-E solution exhibited higher rates for TSF compared with WP (Fig. 4). The difference in TSF is primarily attributable to CHO flux for the CHO-E beverage (Table 2).

Table  2.   Na+, K+, and CHO fluxes in intestinal segments studied Intestinal Segment Na+ Flux K+ Flux CHO Flux WP CHO-E WP CHO-E WP CHO-E 0-25-cm 0.01 ± 0.47   1.04 ± 0.36  0.09 ± 0.02  0.0 ± 0.03  6.0 ± 1.1  25-50-cm 0.19 ± 0.44  1.22 ± 0.33 0.03 ± 0.01  0.12 ± 0.03* 3.3 ± 0.8  50-75-cm 1.99 ± 0.65 1.87 ± 0.10 0.05 ± 0.01  0.07 ± 0.01  0.9 ± 0.3 Values are means ± SE. Na+ and K+ fluxes also represent movement of Cl with each respective cation. Units are meq · cm1 · h1 for Na+ and K+, and mmol · cm1 · h1 for CHO. Negative values represent secretion. * Significantly different from WP. Significantly different from 0- to 25-cm segment.

The osmolality of the WP increased significantly (P < 0.05) at each sampling site in the intestine as it moved distally (Fig. 5), which was reflected in [Na⁺] and [K⁺] changes (Figs. 6 and 7). These changes were smaller during the CHO-E trial, although significant increases (P < 0.05) were observed for values for this solution in the stomach and proximal sampling site compared with sites 25 cm and beyond. Net fluxes for both Na⁺ and K⁺ are presented in Table 2.

Plasma volume was not significantly different between trials at any time point throughout the experiments (Fig. 8). Plasma osmolality increased significantly (P < 0.05) during exercise in both trials but did not change after the 35-min equilibration period, and there was no difference between experiments at any time point except at 45 min, when the WP (292 ± 2 mosmol/kgH₂O) was significantly lower than CHO-E (297 ± 1 mosmol/kgH₂O). Mean plasma osmolality values for the postequilibration period were 293 ± 2 and 295 ± 1 mosmol/kgH₂O for WP and CHO-E, respectively. Plasma [Na⁺] did not change with exercise time, and the values did not differ at any point between beverages (mean values postequilibration = 142.1 ± 0.9 vs. 143.5 ± 0.4 meq/l for WP and CHO-E, respectively). Plasma [K⁺] increased significantly at each time point in both trials (P < 0.05), although no differences were observed between trials at these times, nor were the postequilibration mean values different between trials (4.76 ± 0.06 vs. 4.81 ± 0.06 meq/l for WP and CHO-E, respectively). Plasma glucose concentration was maintained throughout the exercise protocol during ingestion of CHO-E but declined significantly during the WP trial, resulting in significant differences between the two trials for the final 55 min of the experiment (mean values postequilibration = 5.4 ± 0.09 vs. 4.7 ± 1.0 mM for CHO-E and WP, respectively).

DISCUSSION

This is the first study to simultaneously measure intestinal absorption from sequential segments of the proximal small intestine together with GE of beverages consumed orally during exercise. The results provide evidence that fluid absorption from a given solution varies from one segment to another and with the composition of the ingested solution. The following discussion focuses on 1) potential mechanisms to explain this differential segmental absorption (duodenal vs. jejunal), 2) the consequences of bypassing the duodenum in intestinal absorption studies, 3) gastric influences, and 4) the relationship of intestinal absorption to fluid balance, thermoregulation, and energy substrate availability. It is important to note that these experiments were performed by euhydrated, healthy subjects during cycle exercise at 60-65% O_{2 peak}.

Duodenal Absorption (0-25 cm)

Water absorption values obtained in the 0- to 25-cm segment contrast with those previously observed during segmental perfusion of the duodenojejunum. Perfusion of distilled water (infusion rate = 15 ml/min) in the distal duodenum and proximal jejunum during both rest and exercise (70% O_{2 peak}) in two previous investigations (20, 23) produced significantly lower fluid absorption rates (~8-9 ml · cm¹ · h¹) compared with a CHO-E solution similar to that used in this investigation (~12-13 ml · cm¹ · h¹). In the present study, water flux in the duodenum was 30 vs. 15 ml · cm¹ · h¹ for WP and CHO-E, respectively. This discrepancy is attributable to the intestinal segment studied. In the previous reports, the test segment under study consisted primarily of the jejunum. In this, and one other recent study (44), water absorption was greater when the duodenum was studied. Reitemeier et al. (42) also observed rapid rates of water absorption (50% in 3 min and 67% in 5 min) in the distal duodenum/proximal jejunum after infusion of labeled water (isotopic tracer method using D₂O) into the duodenum. These rates closely resemble those of the present study (for WP, ~60% total volume absorbed in 0- to 25-cm segment). In addition, Shi et al. (49) reported water flux for a hypotonic solution exhibited the highest water absorption rates, followed by isotonic and hypertonic solutions, when water flux in the duodenum and jejunum was combined. In contrast, Santangelo and Krejs (45) perfused the stomach with water at 22 ml/min and examined water absorption in a 50-cm mixing segment, 70-cm jejunal test segment, and 70-cm ileal test segment. Their results indicated higher water absorption in the jejunum (7.2 ml · cm¹ · h¹) vs. mixing segment [3.8 ml · cm¹ · h¹ (presumably duodenum)] and the ileum (2.0 ml · cm¹ · h¹). However, because their 50-cm mixing segment ended at the ligament of Treitz (start of the jejunum), and given that the duodenum is only ~25 cm long (10, 33), the other 25 cm of this segment must have been in the stomach. Thus mixing segment water absorption actually occurred in ~25 cm rather than 50 cm, elevating water absorption in the duodenum to ~7.6 ml · cm¹ · h¹, which is approximately the same as that found in the jejunum.

Higher water absorption values in this study compared with those in others may also reflect high GE rates (~18-19 ml/min), which served as the infusion rate when water flux was calculated (52). Most investigators who use segmental perfusion employ infusion rates of 10-15 ml/min.

Jejunal Absorption (25-50 cm and 50-75 cm)

By the final 25-cm segment (50-75 cm), 66% of the WP and 53% of the CHO-E beverage were absorbed, which presumably reduced the flow rate of each to this segment and explains the significant decline in fluid absorption compared with their highest values (0- to 25-cm segment). Furthermore, the osmotic gradient basically disappeared by the end of this segment for the WP beverage (WP = ~260 mosmol/kgH₂O; Fig. 5). There were no differences between the two beverages for fluid absorption in this segment. In the entire 75-cm segment, ~75% of the ingested volume of each solution (minus volume withdrawn for sampling) was absorbed with no differences in total water absorption between solutions.

Using the isotopic tracer technique (D₂O), Davis et al. (12) also observed similar rates of fluid replacement between an isotonic CHO-E beverage and distilled water during cycle exercise in the heat (2 h at 75% O_{2 peak}; 27°C) . In contrast, Leiper et al. (30) and Davis (11) found that isotonic CHO-E solutions were absorbed more readily than water in resting subjects by using this method. It is important to note that isotopic tracer studies do not examine net flux of fluid, only unidirectional flux from intestinal lumen to blood.

It is apparent that had the jejunum been studied exclusively, and the duodenum bypassed, the present results would have been misleading. Interpretation of absorptive efficacy of the beverages (on the basis of only jejunal data) would have favored the CHO-E beverage. However, when duodenal data are considered, fluid absorption was greater for the WP, and thus overall absorption of each beverage indicates no distinct advantage for either.

Gastric Influence

Fluid Homeostasis, Thermoregulation, and Substrate Availability

Plasma glucose concentrations were significantly lower throughout exercise in the WP compared with the CHO-E trial. Absorption of CHO by all segments of the intestine in the CHO-E trial promoted maintenance of plasma glucose concentrations throughout the experiment. Eighty-eight grams of CHO were absorbed within the 75-cm segment studied, or ~1 g/min. This rate meets the maximal blood glucose oxidation rate during prolonged exercise (25). Although performance was not measured in these experiments, ingestion of CHO-containing solutions during prolonged exercise can improve endurance and enhance performance probably through the maintenance of plasma glucose concentration (5-7, 12, 36, 37).

In summary, significant differences exist in intestinal absorption between solutions and among segments in the proximal small intestine. In the duodenum, fluid from the WP beverage is absorbed quickly by movement down a large osmotic gradient despite the possibility of electrolyte secretion. In the jejunum, fluid absorption from the WP beverage is significantly reduced compared with the CHO-E solution, which stimulates greater solute flux, thus promoting greater fluid absorption. Furthermore, these data argue for the inclusion of the duodenum in segmental perfusion studies designed to evaluate the efficacy of oral rehydration solutions because this segment of the intestine plays such a crucial role in fluid absorption. Finally, conclusions drawn from studies that only examine jejunal absorption of a beverage may be misleading when extrapolated to assess the overall effect on fluid homeostasis with oral ingestion of the same beverage.

ACKNOWLEDGEMENTS

The authors thank the subjects for participating in the experiments, Joan Seye for assistance in manuscript preparation, and Pat Johnston, Shawn Fleck, Kay Pals, and Tonya Brueggeman for technical assistance.

FOOTNOTES

Address for reprint requests: C. V. Gisolfi, Dept. of Exercise Science, N414 Field House, Univ. of Iowa, Iowa City, IA 52242-1111.

Received 25 November 1996; accepted in final form 18 March 1997.

REFERENCES