Intestinal water absorption solutions in humans
from select carbohydrate
C. V. GISOLFI, R. W. SUMMERS, H. P. SCHEDL, AND T. L. BLEILER Departments of Exercise Science, Physiology and Biophysics, and Internal Medicine, University of Iowa Colleges of Liberal Arts, Medicine, and the Medical Service, Department of Veterans Affairs Medical Center, Iowa City, Iowa 52242 GISOLFI,C. V., R. W. SUMMERS,H. P. SCHEDL,AND T. L. BLEILER. Irztestinal water absorption from select carbohydrate solutions in humans. J. Appl. Physiol. 73(5): 2142-2150,1992.Eight men positioned a triple-lumen tube in the duodenojejunum. By useof segmentalperfusion, 2,4,6, or 8% solutionsof glucose(111-444 mM), sucrose(55-233 mM), a maltodextrin [17-67 mM, avg. chain length = 7 glucoseunits (7G)], or a corn syrup solid [4O-160 mM, avg. chain length = 3 glucoseunits (3G)] were perfused at 15 ml/min for 70 min after a 3O-min equilibration period. All solutions were made isotonic with NaCl, except 6 and 8% glucosesolutions, which were hypertonic. An isotonic NaCl solution wasperfusedascontrol. Water absorption (range: 9-15 ml. h-l cm-l) did not differ for the 2, 4, and 6% CHO solutions but was greater (P < 0.05) than absorption from control (3.0 t 2.2 ml 3h-l 9cm-‘). The 8% glucose and 3G solutions reduced (P < 0.05) net water flux compared with their 2, 4, and 6% solutions, but 8% sucroseand 8% 7C solutionspromoted water absorption equivalent to lower CHO concentrations. Water absorption wasindependent of [Na+] in the original solution. In the test segment,1) Na’ flux correlated with net water flux (r = 0.72, P c MU), K+ (r = 0.78, P < O.Ol), and [Na+] (r = 0.68, P < 0.001); 2) Na+ absorption occurred at luminal [Na+] aslow as 50 mM; 3) glucosetransport increased linearly over the luminal concentration range of 40-180 mM; and 4) net water flux wassimilar over a range of glucose-to-Na+ concentration ratios of 0.4:l to 3.5:1.We concludethat 1) water absorption is independent of CHO type up to a concentration of 6% for isocaloric solutions, and 2) increasing CHO concentration up to 8% can significantly reduce water absorption for solutions containing glucoseand G3 but not G7 or sucrose. l
sodium ion absorption; segmentalperfusion; oral rehydration therapy DURINGPR~LONGEDEXERCISE, fatigueisassociatedwith
dehydration, hyperthermia, hypoglycemia, and muscle glycogen depletion (2,7,36). In addition, during ultraendurance events, some athletes suffer hyponatremia (14, 32). To promote normal circulatory function, to avoid thermal injury, and to enhance performance, fluids must be ingested during exercise to replace water and salt lost in sweat and to provide an exogenous source of energy (22, 28, 30). These fluids must be rapidly emptied from the stomach and absorbed from the intestine to maintain adequate hydration.
among subjects, and the CHO composition and concentration necessary to maximize water, salt, and hexose absorption is unknown. Fructose stimulates 66-100% as much net water and Na+ absorption as glucose [water absorption is expressed in ml/mmol of CHO absorbed (11)], but fructose absorption is limited in humans (38), and ingestion of fructose solutions can result in gastrointestinal distress (15). Glucose stimulates both active and passive Na+ absorption and K+ secretion (11); fructose stimulates K+ absorption (11). Sucrose inhibits water absorption (31, 35, 49), whereas maltodextrins have been reported to maximize water absorption (20,21,42). With regard to net glucose flux, 1) sucrose has been reported to enhance (46) or retard (21) glucose absorption, 2) maltose confers a kinetic advantage over glucose (5,4l), and 3) glucose derived from maltotriose or a glucose oligomer mixture is absorbed significantly faster than from free glucose (8, 21). The purpose of this study was to compare water absorption from solutions of glucose, sucrose, a maltodextrin, and a corn syrup solid perfused through the duodenojejunum of human subjects. The specific questions ad-
dressed were 1) Does the form of CHO used to formulate a fluid replacement beverage influence water absorption? and 2) At what CHO concentration does water absorption decline? METHODS Subjects. Eight male volunteers aged 25.3 t 1.4 (SE) yr,
weight 75.2 t 2.4 kg, and height 179.5 t 2.2 cm served as subjects in this study. The protocol was approved by our institutional committee on the use of human subjects in research, and each subject provided signed informed consent. Subjects fasted 12 h before passing the multilumen tube, and depending on the duration of the experiment, a maximum of five and a minimum of two solu-
tions were perfused on a given day. Solutions. Seventeen solutions were tested in random order. These included 2,4,6, and 8% solutions of glucose (111-444 mM), sucrose (55-233 mM), a maltodextrin (17-67 mM), a corn syrup solid (40-160 mM), and iso-
tonic NaCl as the control. All solutions were made isotonic by the addition of NaCl except the 6 and 8% glucose
Because virtually no absorption occurs in the stomach (44), hydration and exogenous energy supply is depen-
solutions, which were hypertonic. All solutions contained 1 mg/ml polyethylene glycol as a nonabsorbable marker
dent on carbohydrate from the intestine.
for determination of water flux. The maltodextrins had a dextrose equivalent (DE) of 15, whereas the corn syrup
2142
(CHO), water, and salt absorption Absorption
is markedly
variable
0161-7567192 $2.00 Copyright 0 1992 the American Physiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
INTESTINAL
solids had a DE of 36. The DE value provides an index of the degree of starch polymer hydrolysis; lOO/DE gives the average chain length of the polymer. Thus the maltodextrins had an average chain length of seven glucose units, whereas the corn syrup solids had an average chain length of three glucose units. Thus the maltodextrins are designated G7 and the corn syrup solids are designated G3, as indications of their mean chain lengths. With regard to the frequency distributions of different chain lengths, G3 contained 6% glucose, 53.3% of 2, 3, and 4 glucose units and 22% of 310 glucose units. The remaining percentage contained 5-10 glucose units. G7 contained 1.3% glucose, ~5% each of 2,3,4,5, and 8 glucose units, 18% of 6 and 7 glucose units, 50% of >lO glucose units, and the remaining 5% contained 9 and 10 glucose units. At equal molar concentrations, G7 and G3 solutions provide more glucose moieties for absorption and metabolism than does the free glucose solution. In contrast, sucrose has fewer glucose moieties per gram. Technique. Construction of the multilumen catheter and the segmental perfusion technique have been explained in detail elsewhere (16). Briefly, the multilumen catheter was 150 cm long and consisted of four lumens, each with a 2-mm diameter (Arndorfer Medical Specialties, Greendale, WI). A rubber bag containing 1.5 ml of mercury was sutured to its distal end and enclosed in a balloon. The first lumen of the catheter entered the balloon. The second lumen served as the infusion tube and had a single opening 100 cm from its proximal end. The third and fourth lumens provided sampling sites, each with three openings spaced I cm apart. The middle openings of the proximal and distal sampling sites were 10 and 50 cm distal to the infusion port, respectively. Prot~coZ. Intubations were performed under fluoroscopic guidance in the Digestive Disease Center of the University of Iowa Hospitals and Clinics. Solutions were infused at a rate of 15 ml/min using a Masterflex (Cole Parmer Instrument) pump calibrated immediately before each experiment while the subject remained seated in a chair with back and arm rests. The perfusion protoco1 consisted of a 20-min equilibration period with no perfusion to stabilize plasma volume. At the end of this period, perfusion commenced. The first 30 min served as an equilibration period followed by a 70-min test period. During the test period, fluid samples were drawn from the proximal lumen at 1 ml/min and from the distal lumen by syphonage at IO-min intervals. Blood samples were drawn from an indwelling catheter (18-gauge intracath needle with heparin lock) before the equilibration period and at 15-min intervals during the 70-min test period. Between experiments on the same day, subjects were given lo-20 min between solutions to walk and stretch. This period was followed by the same protocol described above, i.e., 20-min seated rest for plasma volume equilibration followed by 100 min of perfusion (30min equilibration, 70-min test period). Chemical analysis. Polyethylene glycol was measured by the turbidometric assay as modified by Malawer (27). Na+ and K+ were measured by flame photometry (model 943, Instrumentation Laboratories), osmolality was measured by vapor pressure osmometry (model 5500, Wescor), and glucose was measured using Trinder re-
WATER
2143
ABSORPTION I 2% Solution Es 4% Solution 6% Solution q 8% Solution q NaCl
1
I Glucose
t G7
t
I
G3
Sucrose
1 Control
Type of Carbohydrate Water flux in relation to carbohydrate (CHO) concentration grouped by type of CHO (negative values indicate net absorption, positive values net secretion in all figures). With glucose, net water flux decreased with increasing concentration above 4%, and with 8% glucose, water flux did not differ from 0. Water absorption was independent of sucrose concentration. For corn syrup solids with average chain length of 3 glucose units (G3), water absorption decreased progressively with increasing concentration, but this trend was less pronounced for maltodextrins wit,n average chain length of 7 glucose units (G7). * Differs from control NaCl solution (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05). FIG.
1.
agent sured 525), lated
at 505 nm (Sigma no. 315). Hemoglobin was meaby the cyanmethemoglobin method (Sigma no. and percent change in plasma volume was calcuaccording to the method of Dill and Costill (9). Statistical analysis. Net water flux and soluble solute movement were calculated according to Cooper et al. (6). A two-factor repeated-measures analysis of variance (ANOVA) was used to evaluate the effect of time and type of CHO and CHO concentration of the solution on water and solute flux during the test period. An F test for simple effects was used to isolate specific differences. A simple regression analysis of mean Na+ flux on mean water and K+ flux and on Na+ concentration ([Na+]) in the test segment was performed including all solutions. Occasionally, sufficient samples to complete all chemical analyses could not be obtained. When this occurred, if there were no more than two missing values in an experiment, the missing value(s) was replaced with an average of the available values for that experiment. Statistical significance was set at P < 0.05. All values reported are means t SE. RESULTS
water and CHO flux. The order in which solutions were perfused had no systematic effect on any variable (l-way ANOVA, P > 0.05). Also, repeated-measures ANOVA showed no difference in net water, Na+, or K+ flux over time during each experiment. Thus the data were pooled, and only mean values are reported. Figure 1 shows the net water flux data grouped to compare each CHO at the four concentrations. Net water absorption with perfusion of all 2, 4, and 6% CHO solutions was greater (P < 0.05) than with perfusion of the isotonic saline control solution. Net water flux for the 2, 4, and 6% CHO solutions ranged from -11.6 to -13.7, -10.6 to -15.0, and -8.7 to -13.6 ml h-l. cm-‘, respecl
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
1’1’ 40
60
1 80
’
1 100
s
1 120
’
11 140
1 160
m
INTESTINAL
I 180
m
I 200
n
I 220
Test Segment [Glucose], mM FIG.
2. Relationship
concentratiun
between glucose transport
in test segment
during
perfusion
and mean glucose
of 2,4,
6, and 8% gfu-
case solutions. Glucose transport plateaus with hypertonic 6 and 8% solutions. * Differs from 2% solution. t Differs from 4% solution.
tively. Perfusion of 8% sucrose and 8% G7 solutions caused a greater (P < 0.05) absorption of water compared with the control, 8% G3, and 8% glucose solutions. Net water movement during perfusion of hypertonic 8% glucose did not differ from zero. This is attributed to net water secretion in the mixing segment. Net secretion occurred in the mixing segment for both 6 and 8% glucose solutions (50.9 t 13.5 and 58.7 t 11.5 ml 6h-l. cm-‘, respectively). Average net water flux did not differ for sucrose (-13.7 ml h-l cm-l) and G7 (-11.0 ml h-l cm-l) solutions (Fig. 1). All sucrose and G7 solutions produced significantly greater water absorption than the control solution. For the G3 solutions, net water flux was greater (P < 0.05) than control during perfusion of the 2,4, and 6% solutions, and the 2 and 4% solutions produced greater water absorption than the 8% solution. For glucose, the 2, 4, and 6% solutions produced greater (P c 0.05) water absorption than the control and the 8% glucose solution. Net glucose flux measured from the 2, 4, 6, and 8% glucose solutions was -1.5 + 0.5, -2.4 t -0.4, -4.0 t 0.3, and -3.2 -t 0.6 mmol* h-l. cm-‘, respectively, i.e., increased linearly from the 2 to the 6% solution, then plateaued when the concentration was increased to 8% (Fig. 2). The percent glucose absorbed from the 2, 4, 6, and 8% solutions was 75.5 t 6.2, 72.0 t 13.2, 83.1 t 9.2, and 61.2 k 7.7%, respectively. IV& ftux. [Na+] values in the original solution and at the proximal and distal sampling sites are shown in Table 1. [Na+] in the original solution ranged from 0 to 155 mM and did not influence rates of water absorption. [Na+] at the proximal sampling site ranged from 32 to 147 mM, and at the distal site it ranged from 59 to 148 mM. Net water flux correlated with net Na+ flux (Fig. 3, r = 0.72, P < 0.01) but less well with [Na+] in the test segment (r = 0.23, P = 0.02, data not shown). However, net Na+ flux correlated significantly with [Na+3 in the test segment, and net absorption occurred at a [Na+3 as low as 50 mM (Fig. 4). Na+ absorption was greater with all 2% solutions than with the isotonic saline control solution, but there was no difference in Na’ flux among any of the 2% solutions (Fig. 5). Naf flux was greater with the l
l
l
l
4% G7 solution than with the 4% glucose. Within the 4%
WATER
ABSORPTION
solutions, only the G7 solution differed from control. During perfusion of 6% glucose, Na+ movement did not differ from zero, but perfusing the 8% glucose solution resulted in net Na+ secretion, which was significantly different from the control solution. All other 6% solutions were associated with net Na+ absorption. The 8% G7 solution resulted in greater (P < 0.05) Na+ absorption compared with the other 8% CHO types. Na’ flux produced by the 2 and 8% glucose solutions was different (P < 0.05) from those obtained during perfusion of the control solution (Fig. 5). Na+ flux for the 2, 4, and 6% glucose solutions were all significantly different from the 8% solution. Net Na+ flux was the same for the 4 and 6% glucose solutions. There was no difference in net Na+ flux associated with percent CHO in solution during perfusion of the G7 CHO type, but the 2 and 4% solutions produced significantly greater Na+ absorption than control. Net Na’ flux during perfusion of G3 solutions was greater (P < 0.05) than control only for the 2% concentration. Na+ absorption was higher (P < 0.05) during perfusion of the 2 and 4% G3 solutions compared with the 8% solution. Na+ absorption during perfusion of sucrose solutions was greater (P < 0.05) than control only for the 2% concentration.
Na+ absorption was
higher (P < 0.05) during perfusion of 2 and 4% sucrose compared with 8% sucrose. Na+ absorption was also higher for the 2% sucrose solution compared with the 6% sucrose solution. K+ fhx. Perfusion of the control solution caused a net secretion of K+ (Fig. 6). K+ flux was independent of CHO types for 2 or 4% solutions. K+ absorption during perfusion of 2% G3 and 2% sucrose was greater (P < 0.05) than during perfusion of the control solution. Perfusion of all 4% solutions except 4% G3 resulted in significantly (P < 0.05) greater K+ absorption than during perfusion of the control solution. Perfusion of all 6% solutions produced a K+ flux that was not significantly different from the control solution, but the G7 and sucrose solutions were different (P < 0.05) from glucose. Of the 8% solutions, only glucose produced a K+ flux differing from (P < 0.05) the control solution. All other 8% CHO solutions resulted in K+ flux values different (P < 0.05) from the 8% glucose solution. There were no differences in Kf flux associated with different percentages of the G7 or sucrose solutions. K+ flux did not differ from zero with perfusion of 2 and 4% solutions of glucose and G3, whereas the 6 and 8% solutions of glucose and 8% G3 caused a net K+ secretion across the test segment. Net K+ flux correlated significantly with net Na+ flux regardless of CHO concentration (y = 0.04x + 0.29, r = 0.78, P < 0.01). Plasma data. Plasma Na+, K+, and osmolality values did not change (P > 0.05) over time, and therefore they were pooled to yield a single mean value for statistical comparison between groups. When these means were compared, there were no differences between the different forms of CHO or between the different concentrations studied. Table 2 shows the plasma glucose data collected over time. For all CHO solutions, except the 8% glucose solution, there was an increase (P < 0.05) in
plasma glucose by 15 min after the start of perfusion. This elevation remained through 60 min for all solutions except as follows: G3 after 15 min, 4% G7 after 30 min,
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
TABLE 1. Cuncentrution
of carbohydrate
INTESTINAL
WATER
and Nu+ and osmolulity
Carbohydrate
Na’ Concentration, Molarity, mM
%
TYPe Control
0
Glucose
2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
G7
G3
Sucrose
Solution
222 333 444 17 33 50 67 40 80 120 160 58 117 175 233
147.5tl.9* 103.2tl.8
26.0t2.0 7.2~17.1
48.923.5” 42.3t5.9* 32.0t3.9*
0.1,to.o 144.9tll.3 129.6t2.9
different carbohydrate solutions Osmolality,
Distal
Solution
147.6*1.1* 132.8+1.3
276.0-r-1.8
Distal
284.2-t-3.0 277.3k5.6
257.7tl.9 272.5tl.8 275.4tl.9
268.4rt6.9 343.1t2.2 460.4t2.2 288.6t2.0
292.0t2.4* 316.326.5” 297.9t4.1*
275.1+4.6*-/-
28l.Ot3.2
316.644.1*
310.2+6.8
278.1t2.5 288.8+1.2
320.5t5.5” 338.7&12.5*
282.3~11.4
289.725.0
299.843.5 302.4t2.6” 314,9t4*3* 279.9k3.1
301.5+3.l*t 292.4-t4.4*t
99.4+5.3*-F
286.8t2.7 279.8k2.5 281.5~2.8 280.4t2.1 283.5t2.1 272.4ztl.6
311.8+4.0* 323.7&5.6* 288.0t3.3 294.6~13.5 302.9-+8.0*
281.8k3.4 286.8t2.0*
293.1~1.5”~
73.1+5.2*t
284,2+1.5
290.5t3.2”
294.8+2.0*
128.0t8.9 115.2-t-4,7*
99.6t6.9 99.5+2,3*
mosmol/kgH,O
Proximal
276.824.3
84.1+_3.2*t 61.4t5.2 58.7t-2.8”t
120.0+20.7* 122.8t4.0
103.8tl0.4
of the
mM
Proximal
155.5kl.l 90.9a.1
111
2145
ABSORPTION
87.7t3.7 96.1+3.5*
111.8tl.l 111.8104.8 86.9~16.7
125.2tl.9 103.2t3.8
129.3t5.8 109.7dz4.9
84.424.3 55.0~~4.8 123.5-tl.6 82.4t6.4 44.3-t-2.1 16.Ozkl.4
74.6t 10.6 68.7k4.5 122.lk6.5 90.524.3 64.9t2.5* 43.5+3.6*
107.9kl6.3 86.8+7.5*t 140.5+3.2*t 117.4t-2.7*-f
275.0tl.O
277.7-t-1.9*j290.3t2.1
293.5+3.l*t
Values are means + SE of values pooled over the test period. G7, maltodextrans with average chain length of 7 glucose units; G3, corn syrup solids with average chain length of 3 glucose units. * Differs from original solution (P < 0.05). t Differs from proximal (P < 0.05). 2-
-4
0
y=O.lOx+O.24 r 2=O.52 r =0.72 p < 41
-
-4
1
I
I
I
-30
-25
-20
-15
I -10
I -5
I
I
I
0
2-
y=-0.02x+1.10 r 2=o.47
-
I
1
I
15
25
50
Net WaterFlux, ml r h-1 : c$‘l 3. Linear regression of net Na’ flux on net water flux. Each value represents mean of net water flux plotted against mean of net Na+ flux for given solution during last 60 min of perfusion. Fluxes of Na+ and water are directly correlated. FIG.
and 8% sucrose after 45 min. There was no change (P > 0.05) in plasma glucose concentration during perfusion of the isotonic NaCl control solution. Percent change in plasma volume. Values discussed below are mean percent changes in plasma volume (Table 3). Perfusion of the control solution did not significantly change plasma volume. Both 2% glucose and 2% sucrose increased (P < 0.05) plasma volume 9.3 and 9.4%, respectively. The average increase in plasma volume for the 2 and 4% solutions was 7.0 and 3.2%, respectively. Perfusion of 6 and 8% glucose resulted in a significant (P < 0.05) decrease (5.2 and 4.9%, respectively) in plasma volume compared with baseline, the control solution, and all other 6 and 8% solutions, respectively. There were no other differences among 6 and 8% solutions and the control solution. Perfusing the Gi’ and G3 CHO solutions caused a mean increase in plasma volume of 5.5 and 4.1%, respectively, but there were no differences among solutions with different percentages of CHO; these solutions were not different from the control. The 2% sucrose increased (P c 0.05) plasma volume by 9.4%, which was 7.8% above con-
I
Test legrnllt
I
I
I
1
150
175
[Na~5rnIVl
FIG. 4. Linear regression of net Na+ flux on [Na+] in test segment. Each value represents mean of net Na+ flux for given solution during last 60 min of perfusion plotted against mean ENa+] in test segment over same time period. Naf absorption (negative values) increases with increasing [Na+]. Note that absorption occurs at [Na+] as low as 50 mM.
trol, 6.4% above 6% sucrose, and 5.4% more than 8% suciose. The 2% sucrose did not elevate plasma volume significantly more than 4% sucrose. DISCUSSION
Wuter absorption. The first question addressed by this study was “Does the concentration and form of CHO used to formulate a fluid replacement beverage influence water absorption ?” The answer is no, up to a concentration of 6% (Fig. 1). These solutions produced significantly greater water absorption than the control isotonic saline solution, supporting the concept that the addition of CHO to an electrolyte solution enhances water transport (11, 43). The second question addressed was “At what CHO concentration does water absorption significantly decline ?” The answer to this question depends on the form of CHO used to formulate the solution. Water absorption was not significantly reduced by increasing CHO concentration up to 8% for solutions containing G7 or sucrose, but an 8% glucose solution prevented water
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
2146
HUMAN
INTESTINAL
WATER
ABSORPTION
TABLE
2. Plasma gk~ose concentration during perfusion
Carbohydrate
L t @c I 1* * Ij -2.0 GlucoseG?
*I”I5 I G3
Sucrose
I
Control
Type
%
0 min
15 min
30 min
45 min
60 min
Control Glucose
0 2 4 6 8 2 4 6
4.9Iko.3 4.OkO.3 4.6t0.3 4.0t0.5 4.7t0.7 4.0t0.5 4.6t0.5 4.5t0.2 4.2t0.3 4.6t0.4 4.4t0.3 5.2t0.6 4.2kO.5 4.OkO.4 4.4t0.5 4.8kO.2 4.lkO.8
5.ltO.l 5.7t0.2* 6.5tO.5* 6.0t0.4* 6.O-tOt9 5.7t0.6* 6.O-tO.5* 6.6&O-4* 6.9t0.7* 6.420.4* 6.7&0.6* 7.ltO.5* 6.5&0.5* 5.1*0.2* 6.5*0.3* 6.9kO.4” 6.1+0.7*
5.ltO.l 6.6t0.7* 5.7*0.5*p 5,1Ik0.4* 57~10.6 6.0+0.5* 5.7t0.4* 6,7+0.4*
5.1t0.2 5.5+0.3* 5.6-+0.4**$ S.Ot0.3*f 521~0.6 6.ltO.6* 5.2+0.4’f 6.0+0.4*I-$ 5.4*0,4*-! 6.lt0.4* 6.2t0.6* 5.7t0.6 5.7t0.5” 5.4t0.2* 6.lt0.4” 5.8+0.5*-f* 5.5+0.6*
5.0&O. 1 5.9t0.4” 5.7t0.4* 5.0t0.3* 5.0t0.5 5.9t0.5* 5.1t0.5 5.8+0,4*
G7
Type of Carbohydrate 5. Na+ flux in relation to CHO concentration grouped by type of CHO in each solution. As CHU concentration is increased, Na+ flux FIG.
in accord with progressive decrease in [Na+] (see Table 1). Note parallelism between net water flux (Fig. 1) and net Na+ flux (this figure) for all CHO solutions except sucrose. Na+ flux did not differ from 0 with 6% glucose, and net secretion occurred with 8% glucose.
8 G3
decreased
Sucrose
* Differs from control (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05). 7 Differs from 6% solution of same CHO type (P < 0.05).
@ Differs
from
4% solution
of same CHO
type
Glucose Concn, mM
(P -K 0.05).
2 4 6 8 2 4 6 8
5.8+0.4*t 6.1+0.3* 6.3t0.6” 6.OkO.31 6.4t0.6* 5.4t0.3* 6.4t0.5* 6.6t0.4* 5.5t0.8
t t
t
5.4+0.2*-j6.lt0.5* 6.0+0.4*-l5.1+0.4t$ 5.8t0.6* 5.3*0.2* 5.8kO.2”‘r 5.7t0.3”t$ 5.ft0.4
Values are means k SE. * Differs from 0 min (P < 0.05). t Differs
absorption (small net secretion), and water absorption from the 8% G3 solution was significantly less than water absorption from 8% solutions of sucrose and G7. This latter point is of considerable interest because others (21) have shown significantly
greater glucose absorption
from short-chain glucose oligomers than from glucose monomer and sucrose, suggesting that water absorption (not reported) would also be greater. This discrepancy may be related to CHO concentration of the solutions studied. Jones et al. (21) compared isotonic-isocaloric sugar-saline solutions with 140 mM glucose. However, even at this concentration
(140 mM), if perfusing maltotriose
or a glucose oligomer mixture produces greater glucose absorption than glucose monomer or sucrose, these solutions would also be expected to produce greater water absorption. However, we found no difference in water absorption from the 2% solutions of sucrose and G3; the latter might be considered comparable to the oligomer mixture of Jones et al. In addition to concentration and form of CHO, the important characteristics of fluid replacement beverages include [Na+] and osmolality. Except for the 6 and 8% glucose solutions, osmolality was controlled in the present studies by making the solutions isotonic (Table 1). 0.12 I
I
2%Solution
from
15 min (P < 0.05).
I Gfucose
I G7
I
I G3
Sucrose
I Control
Type of Carbohydrate FIG
6. K’ flux in relation
CHO. Net K* flux is absorptive
to CHO
grouped
by type of
at low CHO concentratims
concentration
and tends
toward secretion or near 0 flux at higher CHO concentrations. * Differs from control (P x 0.05). t Differs from 6% solution of same CHO type (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05).
30 min (P < 0.05).
movement. Na+ absorption also correlated significantly with K+ absorption (n = 0.78, P < O.Ol), and, ultimately,
net water flux follows net solute flux. However, the observation that water absorption correlated poorly with [Naf3 in the test segment (r = 0.23) suggests that CHO absorption played the dominant role in determining water movement in this investigation. Fordtran et al. (13) found that, in the jejunum, Na+ absorption is stimulated by the presence of glucose and is markedly influenced by water movement. The primary avenue of Na+ transport was solution drag. Taken together, these 3. Percent change in plasma u&me during perfusion TABLE
Carbohydrate
Change in Plasma Volume,
Type
%
15 min
30 min
Control Glucose
0 2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
0.5tl.8 10.4k2.6 1.8k3.7 -3.8t2.0 -2.3kl.5 5.6k2.6 -o-2+2.2 3.3t2.9 5.9t3.0 3.0tl.3 3.6t2.1 1.41112.4 -0.1tl.8 6.3k2.5 3.7t1.4* 3.621.2 1.9kl.4
1.8t3.1 10*9&l. 1 l.lk3.4 -3.4kl.9 -4.2*1.5* 3.6k3.2 3.lzk2.0 1.9k2.4 3.9k2.9 3.lkl.6 2Jkl.5 2.6~11.4 -1.9kl.9 8.7*2.7* 2.8-tZ.3 4.5kO.8 3.4t2.5
G3
I
from
Adding different amounts of NaCl to the original solutions to make them isotonic did not influence net water absorption. On the other hand, net water flux correlated significantly with net Na+ flux (Fig. 3), indicating that Na+ movement is an important determinant of water
G7
-0.08 L
$ Differs
Sucrose
45 min 2.3k2.5 4.6&1.3?$ 1.9k3.2 -6.5+2.2* -5.8+1.2*-f5.414.2 1.5-t-2.2 6.3-t-2,1* 7.9*3.0* 5.0t2.2* 2.822.6 -2.7k2.41 -1.021.5 11.7+3.0* 6.7t-1.4*$ 3.1kl.3 3.4t2.3
Values are means t_ SE. * Differs from preperfusion 0.05). 7 Differs from 45 min (P < 0.05). $ Differs from 0.05). Q Differs from 15 min (P < 0.05).
% 60 min 2.0t2.0 9.5+1.7$ -1.lt2.9 -7.1-t-2.3* -7.2*1.6*? 5.Ok3.6 3.4t2.1 6.3&1.9* 7.6t0.7* 6.3t2.1* 7.2*1.9*$ 2.1+0.6$ 0.422.3 10.8t2.4* 7.0&1.6*t$ 0.6k2.7 7.2+2.8*-j(0 min) 30 min
(P < (P
from select carbohydrate
C. V. GISOLFI, R. W. SUMMERS, H. P. SCHEDL, AND T. L. BLEILER Departments of Exercise Science, Physiology and Biophysics, and Internal Medicine, University of Iowa Colleges of Liberal Arts, Medicine, and the Medical Service, Department of Veterans Affairs Medical Center, Iowa City, Iowa 52242 GISOLFI,C. V., R. W. SUMMERS,H. P. SCHEDL,AND T. L. BLEILER. Irztestinal water absorption from select carbohydrate solutions in humans. J. Appl. Physiol. 73(5): 2142-2150,1992.Eight men positioned a triple-lumen tube in the duodenojejunum. By useof segmentalperfusion, 2,4,6, or 8% solutionsof glucose(111-444 mM), sucrose(55-233 mM), a maltodextrin [17-67 mM, avg. chain length = 7 glucoseunits (7G)], or a corn syrup solid [4O-160 mM, avg. chain length = 3 glucoseunits (3G)] were perfused at 15 ml/min for 70 min after a 3O-min equilibration period. All solutions were made isotonic with NaCl, except 6 and 8% glucosesolutions, which were hypertonic. An isotonic NaCl solution wasperfusedascontrol. Water absorption (range: 9-15 ml. h-l cm-l) did not differ for the 2, 4, and 6% CHO solutions but was greater (P < 0.05) than absorption from control (3.0 t 2.2 ml 3h-l 9cm-‘). The 8% glucose and 3G solutions reduced (P < 0.05) net water flux compared with their 2, 4, and 6% solutions, but 8% sucroseand 8% 7C solutionspromoted water absorption equivalent to lower CHO concentrations. Water absorption wasindependent of [Na+] in the original solution. In the test segment,1) Na’ flux correlated with net water flux (r = 0.72, P c MU), K+ (r = 0.78, P < O.Ol), and [Na+] (r = 0.68, P < 0.001); 2) Na+ absorption occurred at luminal [Na+] aslow as 50 mM; 3) glucosetransport increased linearly over the luminal concentration range of 40-180 mM; and 4) net water flux wassimilar over a range of glucose-to-Na+ concentration ratios of 0.4:l to 3.5:1.We concludethat 1) water absorption is independent of CHO type up to a concentration of 6% for isocaloric solutions, and 2) increasing CHO concentration up to 8% can significantly reduce water absorption for solutions containing glucoseand G3 but not G7 or sucrose. l
sodium ion absorption; segmentalperfusion; oral rehydration therapy DURINGPR~LONGEDEXERCISE, fatigueisassociatedwith
dehydration, hyperthermia, hypoglycemia, and muscle glycogen depletion (2,7,36). In addition, during ultraendurance events, some athletes suffer hyponatremia (14, 32). To promote normal circulatory function, to avoid thermal injury, and to enhance performance, fluids must be ingested during exercise to replace water and salt lost in sweat and to provide an exogenous source of energy (22, 28, 30). These fluids must be rapidly emptied from the stomach and absorbed from the intestine to maintain adequate hydration.
among subjects, and the CHO composition and concentration necessary to maximize water, salt, and hexose absorption is unknown. Fructose stimulates 66-100% as much net water and Na+ absorption as glucose [water absorption is expressed in ml/mmol of CHO absorbed (11)], but fructose absorption is limited in humans (38), and ingestion of fructose solutions can result in gastrointestinal distress (15). Glucose stimulates both active and passive Na+ absorption and K+ secretion (11); fructose stimulates K+ absorption (11). Sucrose inhibits water absorption (31, 35, 49), whereas maltodextrins have been reported to maximize water absorption (20,21,42). With regard to net glucose flux, 1) sucrose has been reported to enhance (46) or retard (21) glucose absorption, 2) maltose confers a kinetic advantage over glucose (5,4l), and 3) glucose derived from maltotriose or a glucose oligomer mixture is absorbed significantly faster than from free glucose (8, 21). The purpose of this study was to compare water absorption from solutions of glucose, sucrose, a maltodextrin, and a corn syrup solid perfused through the duodenojejunum of human subjects. The specific questions ad-
dressed were 1) Does the form of CHO used to formulate a fluid replacement beverage influence water absorption? and 2) At what CHO concentration does water absorption decline? METHODS Subjects. Eight male volunteers aged 25.3 t 1.4 (SE) yr,
weight 75.2 t 2.4 kg, and height 179.5 t 2.2 cm served as subjects in this study. The protocol was approved by our institutional committee on the use of human subjects in research, and each subject provided signed informed consent. Subjects fasted 12 h before passing the multilumen tube, and depending on the duration of the experiment, a maximum of five and a minimum of two solu-
tions were perfused on a given day. Solutions. Seventeen solutions were tested in random order. These included 2,4,6, and 8% solutions of glucose (111-444 mM), sucrose (55-233 mM), a maltodextrin (17-67 mM), a corn syrup solid (40-160 mM), and iso-
tonic NaCl as the control. All solutions were made isotonic by the addition of NaCl except the 6 and 8% glucose
Because virtually no absorption occurs in the stomach (44), hydration and exogenous energy supply is depen-
solutions, which were hypertonic. All solutions contained 1 mg/ml polyethylene glycol as a nonabsorbable marker
dent on carbohydrate from the intestine.
for determination of water flux. The maltodextrins had a dextrose equivalent (DE) of 15, whereas the corn syrup
2142
(CHO), water, and salt absorption Absorption
is markedly
variable
0161-7567192 $2.00 Copyright 0 1992 the American Physiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
INTESTINAL
solids had a DE of 36. The DE value provides an index of the degree of starch polymer hydrolysis; lOO/DE gives the average chain length of the polymer. Thus the maltodextrins had an average chain length of seven glucose units, whereas the corn syrup solids had an average chain length of three glucose units. Thus the maltodextrins are designated G7 and the corn syrup solids are designated G3, as indications of their mean chain lengths. With regard to the frequency distributions of different chain lengths, G3 contained 6% glucose, 53.3% of 2, 3, and 4 glucose units and 22% of 310 glucose units. The remaining percentage contained 5-10 glucose units. G7 contained 1.3% glucose, ~5% each of 2,3,4,5, and 8 glucose units, 18% of 6 and 7 glucose units, 50% of >lO glucose units, and the remaining 5% contained 9 and 10 glucose units. At equal molar concentrations, G7 and G3 solutions provide more glucose moieties for absorption and metabolism than does the free glucose solution. In contrast, sucrose has fewer glucose moieties per gram. Technique. Construction of the multilumen catheter and the segmental perfusion technique have been explained in detail elsewhere (16). Briefly, the multilumen catheter was 150 cm long and consisted of four lumens, each with a 2-mm diameter (Arndorfer Medical Specialties, Greendale, WI). A rubber bag containing 1.5 ml of mercury was sutured to its distal end and enclosed in a balloon. The first lumen of the catheter entered the balloon. The second lumen served as the infusion tube and had a single opening 100 cm from its proximal end. The third and fourth lumens provided sampling sites, each with three openings spaced I cm apart. The middle openings of the proximal and distal sampling sites were 10 and 50 cm distal to the infusion port, respectively. Prot~coZ. Intubations were performed under fluoroscopic guidance in the Digestive Disease Center of the University of Iowa Hospitals and Clinics. Solutions were infused at a rate of 15 ml/min using a Masterflex (Cole Parmer Instrument) pump calibrated immediately before each experiment while the subject remained seated in a chair with back and arm rests. The perfusion protoco1 consisted of a 20-min equilibration period with no perfusion to stabilize plasma volume. At the end of this period, perfusion commenced. The first 30 min served as an equilibration period followed by a 70-min test period. During the test period, fluid samples were drawn from the proximal lumen at 1 ml/min and from the distal lumen by syphonage at IO-min intervals. Blood samples were drawn from an indwelling catheter (18-gauge intracath needle with heparin lock) before the equilibration period and at 15-min intervals during the 70-min test period. Between experiments on the same day, subjects were given lo-20 min between solutions to walk and stretch. This period was followed by the same protocol described above, i.e., 20-min seated rest for plasma volume equilibration followed by 100 min of perfusion (30min equilibration, 70-min test period). Chemical analysis. Polyethylene glycol was measured by the turbidometric assay as modified by Malawer (27). Na+ and K+ were measured by flame photometry (model 943, Instrumentation Laboratories), osmolality was measured by vapor pressure osmometry (model 5500, Wescor), and glucose was measured using Trinder re-
WATER
2143
ABSORPTION I 2% Solution Es 4% Solution 6% Solution q 8% Solution q NaCl
1
I Glucose
t G7
t
I
G3
Sucrose
1 Control
Type of Carbohydrate Water flux in relation to carbohydrate (CHO) concentration grouped by type of CHO (negative values indicate net absorption, positive values net secretion in all figures). With glucose, net water flux decreased with increasing concentration above 4%, and with 8% glucose, water flux did not differ from 0. Water absorption was independent of sucrose concentration. For corn syrup solids with average chain length of 3 glucose units (G3), water absorption decreased progressively with increasing concentration, but this trend was less pronounced for maltodextrins wit,n average chain length of 7 glucose units (G7). * Differs from control NaCl solution (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05). FIG.
1.
agent sured 525), lated
at 505 nm (Sigma no. 315). Hemoglobin was meaby the cyanmethemoglobin method (Sigma no. and percent change in plasma volume was calcuaccording to the method of Dill and Costill (9). Statistical analysis. Net water flux and soluble solute movement were calculated according to Cooper et al. (6). A two-factor repeated-measures analysis of variance (ANOVA) was used to evaluate the effect of time and type of CHO and CHO concentration of the solution on water and solute flux during the test period. An F test for simple effects was used to isolate specific differences. A simple regression analysis of mean Na+ flux on mean water and K+ flux and on Na+ concentration ([Na+]) in the test segment was performed including all solutions. Occasionally, sufficient samples to complete all chemical analyses could not be obtained. When this occurred, if there were no more than two missing values in an experiment, the missing value(s) was replaced with an average of the available values for that experiment. Statistical significance was set at P < 0.05. All values reported are means t SE. RESULTS
water and CHO flux. The order in which solutions were perfused had no systematic effect on any variable (l-way ANOVA, P > 0.05). Also, repeated-measures ANOVA showed no difference in net water, Na+, or K+ flux over time during each experiment. Thus the data were pooled, and only mean values are reported. Figure 1 shows the net water flux data grouped to compare each CHO at the four concentrations. Net water absorption with perfusion of all 2, 4, and 6% CHO solutions was greater (P < 0.05) than with perfusion of the isotonic saline control solution. Net water flux for the 2, 4, and 6% CHO solutions ranged from -11.6 to -13.7, -10.6 to -15.0, and -8.7 to -13.6 ml h-l. cm-‘, respecl
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
1’1’ 40
60
1 80
’
1 100
s
1 120
’
11 140
1 160
m
INTESTINAL
I 180
m
I 200
n
I 220
Test Segment [Glucose], mM FIG.
2. Relationship
concentratiun
between glucose transport
in test segment
during
perfusion
and mean glucose
of 2,4,
6, and 8% gfu-
case solutions. Glucose transport plateaus with hypertonic 6 and 8% solutions. * Differs from 2% solution. t Differs from 4% solution.
tively. Perfusion of 8% sucrose and 8% G7 solutions caused a greater (P < 0.05) absorption of water compared with the control, 8% G3, and 8% glucose solutions. Net water movement during perfusion of hypertonic 8% glucose did not differ from zero. This is attributed to net water secretion in the mixing segment. Net secretion occurred in the mixing segment for both 6 and 8% glucose solutions (50.9 t 13.5 and 58.7 t 11.5 ml 6h-l. cm-‘, respectively). Average net water flux did not differ for sucrose (-13.7 ml h-l cm-l) and G7 (-11.0 ml h-l cm-l) solutions (Fig. 1). All sucrose and G7 solutions produced significantly greater water absorption than the control solution. For the G3 solutions, net water flux was greater (P < 0.05) than control during perfusion of the 2,4, and 6% solutions, and the 2 and 4% solutions produced greater water absorption than the 8% solution. For glucose, the 2, 4, and 6% solutions produced greater (P c 0.05) water absorption than the control and the 8% glucose solution. Net glucose flux measured from the 2, 4, 6, and 8% glucose solutions was -1.5 + 0.5, -2.4 t -0.4, -4.0 t 0.3, and -3.2 -t 0.6 mmol* h-l. cm-‘, respectively, i.e., increased linearly from the 2 to the 6% solution, then plateaued when the concentration was increased to 8% (Fig. 2). The percent glucose absorbed from the 2, 4, 6, and 8% solutions was 75.5 t 6.2, 72.0 t 13.2, 83.1 t 9.2, and 61.2 k 7.7%, respectively. IV& ftux. [Na+] values in the original solution and at the proximal and distal sampling sites are shown in Table 1. [Na+] in the original solution ranged from 0 to 155 mM and did not influence rates of water absorption. [Na+] at the proximal sampling site ranged from 32 to 147 mM, and at the distal site it ranged from 59 to 148 mM. Net water flux correlated with net Na+ flux (Fig. 3, r = 0.72, P < 0.01) but less well with [Na+] in the test segment (r = 0.23, P = 0.02, data not shown). However, net Na+ flux correlated significantly with [Na+3 in the test segment, and net absorption occurred at a [Na+3 as low as 50 mM (Fig. 4). Na+ absorption was greater with all 2% solutions than with the isotonic saline control solution, but there was no difference in Na’ flux among any of the 2% solutions (Fig. 5). Naf flux was greater with the l
l
l
l
4% G7 solution than with the 4% glucose. Within the 4%
WATER
ABSORPTION
solutions, only the G7 solution differed from control. During perfusion of 6% glucose, Na+ movement did not differ from zero, but perfusing the 8% glucose solution resulted in net Na+ secretion, which was significantly different from the control solution. All other 6% solutions were associated with net Na+ absorption. The 8% G7 solution resulted in greater (P < 0.05) Na+ absorption compared with the other 8% CHO types. Na’ flux produced by the 2 and 8% glucose solutions was different (P < 0.05) from those obtained during perfusion of the control solution (Fig. 5). Na+ flux for the 2, 4, and 6% glucose solutions were all significantly different from the 8% solution. Net Na+ flux was the same for the 4 and 6% glucose solutions. There was no difference in net Na+ flux associated with percent CHO in solution during perfusion of the G7 CHO type, but the 2 and 4% solutions produced significantly greater Na+ absorption than control. Net Na’ flux during perfusion of G3 solutions was greater (P < 0.05) than control only for the 2% concentration. Na+ absorption was higher (P < 0.05) during perfusion of the 2 and 4% G3 solutions compared with the 8% solution. Na+ absorption during perfusion of sucrose solutions was greater (P < 0.05) than control only for the 2% concentration.
Na+ absorption was
higher (P < 0.05) during perfusion of 2 and 4% sucrose compared with 8% sucrose. Na+ absorption was also higher for the 2% sucrose solution compared with the 6% sucrose solution. K+ fhx. Perfusion of the control solution caused a net secretion of K+ (Fig. 6). K+ flux was independent of CHO types for 2 or 4% solutions. K+ absorption during perfusion of 2% G3 and 2% sucrose was greater (P < 0.05) than during perfusion of the control solution. Perfusion of all 4% solutions except 4% G3 resulted in significantly (P < 0.05) greater K+ absorption than during perfusion of the control solution. Perfusion of all 6% solutions produced a K+ flux that was not significantly different from the control solution, but the G7 and sucrose solutions were different (P < 0.05) from glucose. Of the 8% solutions, only glucose produced a K+ flux differing from (P < 0.05) the control solution. All other 8% CHO solutions resulted in K+ flux values different (P < 0.05) from the 8% glucose solution. There were no differences in Kf flux associated with different percentages of the G7 or sucrose solutions. K+ flux did not differ from zero with perfusion of 2 and 4% solutions of glucose and G3, whereas the 6 and 8% solutions of glucose and 8% G3 caused a net K+ secretion across the test segment. Net K+ flux correlated significantly with net Na+ flux regardless of CHO concentration (y = 0.04x + 0.29, r = 0.78, P < 0.01). Plasma data. Plasma Na+, K+, and osmolality values did not change (P > 0.05) over time, and therefore they were pooled to yield a single mean value for statistical comparison between groups. When these means were compared, there were no differences between the different forms of CHO or between the different concentrations studied. Table 2 shows the plasma glucose data collected over time. For all CHO solutions, except the 8% glucose solution, there was an increase (P < 0.05) in
plasma glucose by 15 min after the start of perfusion. This elevation remained through 60 min for all solutions except as follows: G3 after 15 min, 4% G7 after 30 min,
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
HUMAN
TABLE 1. Cuncentrution
of carbohydrate
INTESTINAL
WATER
and Nu+ and osmolulity
Carbohydrate
Na’ Concentration, Molarity, mM
%
TYPe Control
0
Glucose
2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
G7
G3
Sucrose
Solution
222 333 444 17 33 50 67 40 80 120 160 58 117 175 233
147.5tl.9* 103.2tl.8
26.0t2.0 7.2~17.1
48.923.5” 42.3t5.9* 32.0t3.9*
0.1,to.o 144.9tll.3 129.6t2.9
different carbohydrate solutions Osmolality,
Distal
Solution
147.6*1.1* 132.8+1.3
276.0-r-1.8
Distal
284.2-t-3.0 277.3k5.6
257.7tl.9 272.5tl.8 275.4tl.9
268.4rt6.9 343.1t2.2 460.4t2.2 288.6t2.0
292.0t2.4* 316.326.5” 297.9t4.1*
275.1+4.6*-/-
28l.Ot3.2
316.644.1*
310.2+6.8
278.1t2.5 288.8+1.2
320.5t5.5” 338.7&12.5*
282.3~11.4
289.725.0
299.843.5 302.4t2.6” 314,9t4*3* 279.9k3.1
301.5+3.l*t 292.4-t4.4*t
99.4+5.3*-F
286.8t2.7 279.8k2.5 281.5~2.8 280.4t2.1 283.5t2.1 272.4ztl.6
311.8+4.0* 323.7&5.6* 288.0t3.3 294.6~13.5 302.9-+8.0*
281.8k3.4 286.8t2.0*
293.1~1.5”~
73.1+5.2*t
284,2+1.5
290.5t3.2”
294.8+2.0*
128.0t8.9 115.2-t-4,7*
99.6t6.9 99.5+2,3*
mosmol/kgH,O
Proximal
276.824.3
84.1+_3.2*t 61.4t5.2 58.7t-2.8”t
120.0+20.7* 122.8t4.0
103.8tl0.4
of the
mM
Proximal
155.5kl.l 90.9a.1
111
2145
ABSORPTION
87.7t3.7 96.1+3.5*
111.8tl.l 111.8104.8 86.9~16.7
125.2tl.9 103.2t3.8
129.3t5.8 109.7dz4.9
84.424.3 55.0~~4.8 123.5-tl.6 82.4t6.4 44.3-t-2.1 16.Ozkl.4
74.6t 10.6 68.7k4.5 122.lk6.5 90.524.3 64.9t2.5* 43.5+3.6*
107.9kl6.3 86.8+7.5*t 140.5+3.2*t 117.4t-2.7*-f
275.0tl.O
277.7-t-1.9*j290.3t2.1
293.5+3.l*t
Values are means + SE of values pooled over the test period. G7, maltodextrans with average chain length of 7 glucose units; G3, corn syrup solids with average chain length of 3 glucose units. * Differs from original solution (P < 0.05). t Differs from proximal (P < 0.05). 2-
-4
0
y=O.lOx+O.24 r 2=O.52 r =0.72 p < 41
-
-4
1
I
I
I
-30
-25
-20
-15
I -10
I -5
I
I
I
0
2-
y=-0.02x+1.10 r 2=o.47
-
I
1
I
15
25
50
Net WaterFlux, ml r h-1 : c$‘l 3. Linear regression of net Na’ flux on net water flux. Each value represents mean of net water flux plotted against mean of net Na+ flux for given solution during last 60 min of perfusion. Fluxes of Na+ and water are directly correlated. FIG.
and 8% sucrose after 45 min. There was no change (P > 0.05) in plasma glucose concentration during perfusion of the isotonic NaCl control solution. Percent change in plasma volume. Values discussed below are mean percent changes in plasma volume (Table 3). Perfusion of the control solution did not significantly change plasma volume. Both 2% glucose and 2% sucrose increased (P < 0.05) plasma volume 9.3 and 9.4%, respectively. The average increase in plasma volume for the 2 and 4% solutions was 7.0 and 3.2%, respectively. Perfusion of 6 and 8% glucose resulted in a significant (P < 0.05) decrease (5.2 and 4.9%, respectively) in plasma volume compared with baseline, the control solution, and all other 6 and 8% solutions, respectively. There were no other differences among 6 and 8% solutions and the control solution. Perfusing the Gi’ and G3 CHO solutions caused a mean increase in plasma volume of 5.5 and 4.1%, respectively, but there were no differences among solutions with different percentages of CHO; these solutions were not different from the control. The 2% sucrose increased (P c 0.05) plasma volume by 9.4%, which was 7.8% above con-
I
Test legrnllt
I
I
I
1
150
175
[Na~5rnIVl
FIG. 4. Linear regression of net Na+ flux on [Na+] in test segment. Each value represents mean of net Na+ flux for given solution during last 60 min of perfusion plotted against mean ENa+] in test segment over same time period. Naf absorption (negative values) increases with increasing [Na+]. Note that absorption occurs at [Na+] as low as 50 mM.
trol, 6.4% above 6% sucrose, and 5.4% more than 8% suciose. The 2% sucrose did not elevate plasma volume significantly more than 4% sucrose. DISCUSSION
Wuter absorption. The first question addressed by this study was “Does the concentration and form of CHO used to formulate a fluid replacement beverage influence water absorption ?” The answer is no, up to a concentration of 6% (Fig. 1). These solutions produced significantly greater water absorption than the control isotonic saline solution, supporting the concept that the addition of CHO to an electrolyte solution enhances water transport (11, 43). The second question addressed was “At what CHO concentration does water absorption significantly decline ?” The answer to this question depends on the form of CHO used to formulate the solution. Water absorption was not significantly reduced by increasing CHO concentration up to 8% for solutions containing G7 or sucrose, but an 8% glucose solution prevented water
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (163.015.154.053) on October 15, 2018. Copyright © 1992 American Physiological Society. All rights reserved.
2146
HUMAN
INTESTINAL
WATER
ABSORPTION
TABLE
2. Plasma gk~ose concentration during perfusion
Carbohydrate
L t @c I 1* * Ij -2.0 GlucoseG?
*I”I5 I G3
Sucrose
I
Control
Type
%
0 min
15 min
30 min
45 min
60 min
Control Glucose
0 2 4 6 8 2 4 6
4.9Iko.3 4.OkO.3 4.6t0.3 4.0t0.5 4.7t0.7 4.0t0.5 4.6t0.5 4.5t0.2 4.2t0.3 4.6t0.4 4.4t0.3 5.2t0.6 4.2kO.5 4.OkO.4 4.4t0.5 4.8kO.2 4.lkO.8
5.ltO.l 5.7t0.2* 6.5tO.5* 6.0t0.4* 6.O-tOt9 5.7t0.6* 6.O-tO.5* 6.6&O-4* 6.9t0.7* 6.420.4* 6.7&0.6* 7.ltO.5* 6.5&0.5* 5.1*0.2* 6.5*0.3* 6.9kO.4” 6.1+0.7*
5.ltO.l 6.6t0.7* 5.7*0.5*p 5,1Ik0.4* 57~10.6 6.0+0.5* 5.7t0.4* 6,7+0.4*
5.1t0.2 5.5+0.3* 5.6-+0.4**$ S.Ot0.3*f 521~0.6 6.ltO.6* 5.2+0.4’f 6.0+0.4*I-$ 5.4*0,4*-! 6.lt0.4* 6.2t0.6* 5.7t0.6 5.7t0.5” 5.4t0.2* 6.lt0.4” 5.8+0.5*-f* 5.5+0.6*
5.0&O. 1 5.9t0.4” 5.7t0.4* 5.0t0.3* 5.0t0.5 5.9t0.5* 5.1t0.5 5.8+0,4*
G7
Type of Carbohydrate 5. Na+ flux in relation to CHO concentration grouped by type of CHO in each solution. As CHU concentration is increased, Na+ flux FIG.
in accord with progressive decrease in [Na+] (see Table 1). Note parallelism between net water flux (Fig. 1) and net Na+ flux (this figure) for all CHO solutions except sucrose. Na+ flux did not differ from 0 with 6% glucose, and net secretion occurred with 8% glucose.
8 G3
decreased
Sucrose
* Differs from control (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05). 7 Differs from 6% solution of same CHO type (P < 0.05).
@ Differs
from
4% solution
of same CHO
type
Glucose Concn, mM
(P -K 0.05).
2 4 6 8 2 4 6 8
5.8+0.4*t 6.1+0.3* 6.3t0.6” 6.OkO.31 6.4t0.6* 5.4t0.3* 6.4t0.5* 6.6t0.4* 5.5t0.8
t t
t
5.4+0.2*-j6.lt0.5* 6.0+0.4*-l5.1+0.4t$ 5.8t0.6* 5.3*0.2* 5.8kO.2”‘r 5.7t0.3”t$ 5.ft0.4
Values are means k SE. * Differs from 0 min (P < 0.05). t Differs
absorption (small net secretion), and water absorption from the 8% G3 solution was significantly less than water absorption from 8% solutions of sucrose and G7. This latter point is of considerable interest because others (21) have shown significantly
greater glucose absorption
from short-chain glucose oligomers than from glucose monomer and sucrose, suggesting that water absorption (not reported) would also be greater. This discrepancy may be related to CHO concentration of the solutions studied. Jones et al. (21) compared isotonic-isocaloric sugar-saline solutions with 140 mM glucose. However, even at this concentration
(140 mM), if perfusing maltotriose
or a glucose oligomer mixture produces greater glucose absorption than glucose monomer or sucrose, these solutions would also be expected to produce greater water absorption. However, we found no difference in water absorption from the 2% solutions of sucrose and G3; the latter might be considered comparable to the oligomer mixture of Jones et al. In addition to concentration and form of CHO, the important characteristics of fluid replacement beverages include [Na+] and osmolality. Except for the 6 and 8% glucose solutions, osmolality was controlled in the present studies by making the solutions isotonic (Table 1). 0.12 I
I
2%Solution
from
15 min (P < 0.05).
I Gfucose
I G7
I
I G3
Sucrose
I Control
Type of Carbohydrate FIG
6. K’ flux in relation
CHO. Net K* flux is absorptive
to CHO
grouped
by type of
at low CHO concentratims
concentration
and tends
toward secretion or near 0 flux at higher CHO concentrations. * Differs from control (P x 0.05). t Differs from 6% solution of same CHO type (P < 0.05). ’ Differs from 8% solution of same CHO type (P < 0.05).
30 min (P < 0.05).
movement. Na+ absorption also correlated significantly with K+ absorption (n = 0.78, P < O.Ol), and, ultimately,
net water flux follows net solute flux. However, the observation that water absorption correlated poorly with [Naf3 in the test segment (r = 0.23) suggests that CHO absorption played the dominant role in determining water movement in this investigation. Fordtran et al. (13) found that, in the jejunum, Na+ absorption is stimulated by the presence of glucose and is markedly influenced by water movement. The primary avenue of Na+ transport was solution drag. Taken together, these 3. Percent change in plasma u&me during perfusion TABLE
Carbohydrate
Change in Plasma Volume,
Type
%
15 min
30 min
Control Glucose
0 2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
0.5tl.8 10.4k2.6 1.8k3.7 -3.8t2.0 -2.3kl.5 5.6k2.6 -o-2+2.2 3.3t2.9 5.9t3.0 3.0tl.3 3.6t2.1 1.41112.4 -0.1tl.8 6.3k2.5 3.7t1.4* 3.621.2 1.9kl.4
1.8t3.1 10*9&l. 1 l.lk3.4 -3.4kl.9 -4.2*1.5* 3.6k3.2 3.lzk2.0 1.9k2.4 3.9k2.9 3.lkl.6 2Jkl.5 2.6~11.4 -1.9kl.9 8.7*2.7* 2.8-tZ.3 4.5kO.8 3.4t2.5
G3
I
from
Adding different amounts of NaCl to the original solutions to make them isotonic did not influence net water absorption. On the other hand, net water flux correlated significantly with net Na+ flux (Fig. 3), indicating that Na+ movement is an important determinant of water
G7
-0.08 L
$ Differs
Sucrose
45 min 2.3k2.5 4.6&1.3?$ 1.9k3.2 -6.5+2.2* -5.8+1.2*-f5.414.2 1.5-t-2.2 6.3-t-2,1* 7.9*3.0* 5.0t2.2* 2.822.6 -2.7k2.41 -1.021.5 11.7+3.0* 6.7t-1.4*$ 3.1kl.3 3.4t2.3
Values are means t_ SE. * Differs from preperfusion 0.05). 7 Differs from 45 min (P < 0.05). $ Differs from 0.05). Q Differs from 15 min (P < 0.05).
% 60 min 2.0t2.0 9.5+1.7$ -1.lt2.9 -7.1-t-2.3* -7.2*1.6*? 5.Ok3.6 3.4t2.1 6.3&1.9* 7.6t0.7* 6.3t2.1* 7.2*1.9*$ 2.1+0.6$ 0.422.3 10.8t2.4* 7.0&1.6*t$ 0.6k2.7 7.2+2.8*-j(0 min) 30 min
(P < (P
