Gut, 1990, 31, 1115-1119 1115 Evaluation of the efficacy of oral rehydration solutions using human whole gut perfusion Gut: first published as 10.1136/gut.31.10.1115 on 1 October 1990. Downloaded from
D D K Rolston, S N Zinzuvadia, V I Mathan
Abstract exist between different parts of the small and Whole gut perfusion in humans was used to large intestines.'3 compare the effect on intestinal water and These problems with methods currently used electrolyte transport of the World Health for evaluating the efficacy of oral rehydration Organisation oral rehydration solution (solu- solutions led us to use human whole gut per- tion II, composition in mmolIl: glucose 111, fusion in vivo in healthy subjects to compare the sodium 90, bicarbonate 30, potassium 20; effect of the World Health Organisation (WHO) 308 mOsm/kg); a hypertonic commercial oral bicarbonate oral rehydration solution, a com- rehydration solution (solution Ill, glucose 188, mercial solution (Peditral, Searle (India) Ltd) sodium 50, bicarbonate 20, potassium and three experimental glucose-electrolyte solu- 20 mmolIl; 335 mOsm/kg); and three experi- tions on water and electrolyte transport. The mental bicarbonate free, hypotonic oral re- composition of the experimental solutions is hydration solutions: solution IV (glucose 111, based on the following observations: (i) seg- sodium 60, potassium 20 mmol/l; 260 mOsm/ mental perfusion studies in healthy Indian sub- kg), solution V (glucose 80, sodium 60, potas- jects have shown that 80 mmol/l glucose stimu- sium 20 mmolIl; 219 mOsm/kg), and solution lates jejunal water absorption maximally'4; VI (glucose 80, sodium 30, potassium (ii) WHO recommends as optimal IIl mmol/l 20 mmolll; 177 mOsm/kg). Perfusion of the glucose in oral rehydration solutions'5; (iii) bicar- intestine with a standard cleansing solution bonate in oral rehydration solutions may not be (solution I, sodium 125, potassium 10, bicar- necessary to correct acidosis'6; and (iv) hypotonic bonate 20, sulphate 40, mannitol 80 mmol/l; glucose-electrolyte solutions result in greater 275 mOsm/kg) confirmed published data on water absorption than isotonic glucose-electro- minimal water and sodium absorption. Experi- lyte solutions.5 mental solution VI produced maximum water During whole gut perfusion these solutions absorption (mean (SE) +1660*0 (29.8) ml/h) were instilled directly into the third part of the http://gut.bmj.com/ significantly greater than solution II (+1195*3 duodenum. In practice, however, oral rehydra- (79-5) ml/h), III (+534.7 (140.3) ml/h), IV tion solutions are taken by mouth. It is possible (+1498.0 (42.7) ml/h), and V (+1327.7 (24.4) that when taken orally the glucose concentration ml/h; p<005). Sodium absorption was signifi- in these solutions greatly decrease by the time the cantly greater with solution II (+97.4 (7.9) solution reaches the distal duodenum because of mmol/h) compared to VI (+43.3 (7.8) mmol/h; dilution with gastric, pancreatic, biliary, and p
TABLE I Composition ofsolutions perfused* (mmolll) ANALYSIS OF SAMPLES AND CALCULATIONS Aliquots from each collection period during the Osmolality Solution Glucose Na' K' HCO3 Cl (mOsm/kg) whole gut perfusions were stored at -20°C until Gut: first published as 10.1136/gut.31.10.1115 on 1 October 1990. Downloaded from analysed for glucose by the glucose oxidase It 125 10 20 35 275 II 111 90 20 30 80 308 method,'9 polyethylene glycol by the turbidi- (WHO-oral rehydration solution) metric method of Hyden,20 sodium and potas- III 188 50 20 20 50 335 (Peditral)t sium by flame photometry, chloride by an EEL IV 111 60 20 - 80 260 chloridometer, and total CO2 in an automated V 80 60 20 - 80 219 VI 80 30 20 - 50 177 Corning 965 CO2 analyser (Corning Ltd, Hal- stead, Essex, UK). Osmolality was measured in a *All solutions contained polyethylene glycol 4000 (2-5 g/l) as a non-absorbable volume marker. tAlso contains S04 40 mmolI and mannitol 80 mmolUl. Wescor vapour pressure osmometer (Wescor tSearle (India) Ltd. Inc, St Logan, USA). Glucose, water, and electrolyte fluxes were calculated using standard formulas.2' Net absorption (+) denotes net centage ofthe mean, was <± 8%). This occurred transfer from the intestinal lumen; net secretion half an hour after no sulphobromophthalein was (-) denotes transfer into the lumen. detected in the rectal effluent, as judged by the absence of a colour change on alkalinising the effluent with saturated sodium hydroxide. STATISTICAL METHODS The subject was perfused with solution I Data are presented as the mean (SE). Differences (Table I) in the evening to clean the intestinal were considered significant at p<0 05 by the lumen of food debris and faecal matter. This Mann-Whitney U test. solution has been shown to result in minimal water and electrolyte fluxes during whole gut perfusion,'8 and is an effective means of cleans- Results ing the intestine of particulate matter. The polyvinyl tube was left in place overnight and the WHOLE GUT PERFUSION subject was permitted only water. The next Water and electrolytefluxes morning the subject was perfused with one of All solutions containing glucose (II-VI) stimu- the glucose-electrolyte solutions (Table I). lated water absorption (Fig 1). Hypotonic experi- mental solutions IV and VI enhanced water absorption to a significantly greater extent than GLUCOSE CONCENTRATION PROFILE OF THE SMALL the bicarbonate WHO solution (solution II; INTESTINE p<005) and solution VI to a greater extent than
Six healthy volunteers were studied to determine solution IV (p<0 05). The commercial hyper- http://gut.bmj.com/ the glucose concentration profile of the small tonic solution (Peditral; solution III) produced intestine after the subject had ingested solutions significantly less water absorption (p<0-01) com- containing glucose. Each subject swallowed four pared to all other solutions containing glucose. polyvinyl tubes glued together and weighted Sodium absorption occurred with all solutions with a mercury bag. These tubes were positioned (Fig 2). Maximal sodium absorption occurred fluoroscopically such that the proximal tube with solution II (WHO), which contained the
opening was in the jejunum immediately distal to on September 23, 2021 by guest. Protected copyright. the ligament of Treitz and the openings of the other three tubes were at 45, 90, and 135 cm, 1800 r t respectively, from the ligament of Treitz. The 4- subject then drank 600 ml ofa glucose solution in 1600 l4: water over a 10 minute period. Solutions con- taining glucose in concentrations of 100, 140, 1400 § 250, and 350 mmol/l were studied. Some 1-2 ml -f 1200 of the ingested solution was aspirated from each c tube and stored at -20°C until analysed for 0 glucose. Three hours later the subject ingested o 1000 another glucose solution and intestinal fluid was n aspirated and stored as described above. a} 800 I. t All reagents used were ofanalytical grade. This study was approved by the Research and z 600 I. I Ethics Committee of the Christian Medical 400j College Hospital, Vellore. 200[
TABLE II Glucose, potassium, chloride, and bicarbonate transport (mmollh) during whole gut ---A- perfusions (mean (SE)) Solution No _ I III IV V VI Solution Glucose K Cl HCO3 Glucose (mmol/l) 0 111 188 111 80 80 I + 4-7(06) - 2-5(1 8) +119(1 6) Glucose:sodium 0 1-2 3-8 1-9 1*3 2 7 II +182.7(0 7) +25-6(1-0) + 89-1(8-8) +31-4(2 8) III +217-1 (4 2) + 17-1 (2-4)* + 32-7 (5-3)* + 9-2 (2-3)* IV +183-9(1 2) +26 2(2-7) + 99.0(16 5) -10 4(2 0)* In=21 7 5 5 5 V +127 9(0-2) +26 3(1 2) +101.0(6-1) -10-4(1-6)* VI +129-5 (0-1) +30 3 (0-4)* + 78-8 (1 8) - 9-1 (3 4)* Figure 1: Net water transport during whole gut perfusion with a lavage solution (I) andglucose-electrolyte solutions (II-IV). + =net absorption; -= net secretion. *p<005, and tp Figure 2: Net sodium 120 transport during whole gut perfusion with a lavage Gut: first published as 10.1136/gut.31.10.1115 on 1 October 1990. Downloaded from solution (I) and glucose- electrolyte solutions (II-VI). 100 *p 20- 4 90 135 180 225 Initial glucose Distance of orifices from incisors (cm) concentration n Figure 3: Glucose concentration profile ofthe small intestine UN I.I I -II* X after ingestion offour glucose solutions with different Solution No I II II IV V VI concentrations; n=5for allpoints, except at 225 cm - that is, cm the small intestine distal to the ligament of Initial sodium in thefirst 135 of 125 90 50 60 60 30 Treitz, where it was not always possible to obtainfluid. Data concconc(mmso/i)( mmol/1) are expressed as mean (SE). n= 21 7 5 6 5 5 from the glucose concentration in the ingested solution (Fig 3). Glucose was almost completely highest initial sodium concentration (90 absorbed in the first 135 cm ofthe small intestine mmol/l). Solution I resulted in minimal sodium of Treitz. absorption. distal to the ligament Potassium and chloride absorption also occurred with all solutions containing glucose. Discussion Bicarbonate absorption occurred with solutions The glucose concentration profiles in the small containing bicarbonate (II and III), whereas obtained after solutions with different bicarbonate secretion occurred when bicar- intestine concentrations of glucose were drunk show that http://gut.bmj.com/ bonate free solutions (IV-VI) were perfused the glucose concentration does not change (Table II). At the end of the study, however, appreciably until the proximal jejunum. This none ofthe subjects was clinically acidotic. observation justifies the use of the whole gut perfusion technique (in which solutions were perfused into the distal duodenum) to compare Other results different solu- The time required for the rectal effluent to the effects of glucose-electrolyte tions, which in practice are taken orally, on water on September 23, 2021 by guest. Protected copyright. become free of sulphobromophthalein after and electrolyte transport. injection into the infusion port (equilibration In attempting to design an oral rehydration period) ranged from 1 hour 5 minutes to 4 hours which could produce maximum water 30 minutes (Table III). solution Rectal flow rate was significantly greater with absorption, we took advantage of data from solution I compared to II, IV, V, and VI human'422 and animal studies56 which showed (p<0 001, Table III). It was least when solution that the glucose concentration, osmolality, and VI was which produced the greatest the glucose:sodium ratio are important deter- perfused minants of intestinal water absorption from a water absorption. glucose-electrolyte solution. The present study supports jejunal segmental perfusion data'423 in that solutions with glucose concentrations of SMALL INTESTINAL GLUCOSE CONCENTRATION in diminished water PROFILE -200 mmol/l resulted After the subject had drunk a solution containing absorption compared to lower glucose concen- glucose the glucose concentration in the trations. It is not surprising that the effect of jejunum was not statistically different glucose on water absorption derived from human proximal and animal jejunal segmental and human whole gut perfusions are similar, because glucose exerts TABLE III Rectal effluentflow rate and time requiredfor its maximum effect on water transport in the sulphobromophthalein (BSP) clearance (mean (SE)) jejunum and is not a potent stimulant of water absorption in the ileum'4 and colon.24 The Rectal effluent Time required No of flow rate for BSP glucose concentration profile in the small intes- Solution subjects (ml/min) clearance (h) tine, after solutions containing glucose were I 21 32-1(2 0) 2-3 (0-2) drunk, confirms that nearly all the ingested II 7 85 (1-8)* 3-2(0-4) glucose is absorbed in the jejunum irrespective of III 5 23-8 (8-9) 1-8 (0 2) IV 6 4-1 (0-9)* 2-6 (0-2) the initial concentration. V 5 4-9(1-3)* 2-4(0-3) In contrast to glucose, there is no parallel in VI 5 2-1(0-9)* 3-1(0-3) sodium transport between human and animal *p whole gut perfusions. Small intestinal perfusions on glucose absorption and a less striking effect on in both rats and humans have shown that sodium water absorption during segmental perfusions.30 absorption usually occurs at concentrations >90 As water is absorbed the flow rate further down Gut: first published as 10.1136/gut.31.10.1115 on 1 October 1990. Downloaded from mmol/l.525 It is possible that this is a peculiarity the intestinal segment decreases and luminal of the segmental perfusion system as here the glucose concentrations tend to remain high role of the colon in sodium homoeostasis is because of the relative concentrating effect. The completely ignored. Indeed, the human colon luminal glucose concentration, in turn, has a has an equilibrium concentration for sodium profound effect on water absorption'43' It is absorption as low as 30 mmol/l and it is probable therefore important to bear in mind that the that the relatively inefficient sodium absorption results of the present study, where a high flow by the jejunum is offset by efficient sodium rate was used to obtain reasonable rectal effluent absorption by the colon. Furthermore, the flow rates, might have been different if a higher flow rate in the whole gut perfusion study different flow rate had been used. could result in a better mixed, but thinner Our data provide evidence that solution VI, unstirred water layer than at the lower flow rates which is a hypotonic bicarbonate free solution used in the segmental perfusions and allow for containing 80 mmolIl glucose, could be the oral more efficient sodium absorption. rehydration solution of choice in combating The glucose:sodium ratio is believed to dehydration due to acute diarrhoea in the tropics. influence water absorption from a glucose solu- The results ofthis study, which are derived from tion containing sodium. It has been suggested intestinal perfusions performed on healthy sub- that glucose:sodium ratios of 2:1 to 2-8:1 pro- jects, should, however, be extrapolated with duce maximum water absorption.526 In the caution to patients with acute diarrhoea because present study, solution VI, which has a glucose: during acute diarrhoea oral rehydration solu- sodium ratio of 2-7, did stimulate water absorp- tions are ingested as boluses and the intestine is tion at a significantly greater rate than the other in a net secretory state. solutions (II to V) which have glucose:sodium The technical assistance of S Jayakumar and the financial support ratios of either <2-1 or >2-8. These solutions, of Searle (India) and Searle (UK) Ltd in carrying out this study is however, have markedly different osmolalities gratefully acknowledged. The Wellcome Research Unit is sup- ported by the Wellcome Trust, London, UK, in association with which could have contributed to the results Medical College, Vellore, India. obtained. It is impossible to determine from the the Christian present study the relative contributions of 1 Islam MR. Can potassium citrate replace sodium bicarbonate and glucose:sodium ratio on net water and potassium chloride of oral rehydration solution? Arch osmolality Dis Child 1985; 60: 852-3. transport. 2 Vesikari T and Isolauri E. Glycine supplemented oral rehydra- One of the reasons for including base precur- tion solutions for diarrhoea. Arch Dis Child 1986; 61: 372-6. sors - for example, bicarbonate, acetate, and 3 Bhargava SK, Sachdev HPS, Das Gupta B, Daral TS, Singh http://gut.bmj.com/ citrate - in oral rehydration solutions is that they HP, Mohan M. Oral rehydration of neonates and young infants with dehydrating diarrhea: comparison of low and stimulate water and sodium absorption. lo27 28 standard content in oral rehydration solutions. J Pediatr The present study suggests that a glucose solu- Gastroenterol Nutr 1984; 3: 500-5. 4 Alam AN, Sarker SA, Molla AM, Rahaman MM, Greenough tion containing bicarbonate (solution II) does not WB III. Hydrolysed wheat-based oral rehydration solution provide an advantage in water absorption com- for acute diarrhoea. Arch Dis Child 1987; 62: 440-4. 5 Rolston DDK, Borodo MM, Kelly MJ, Dawson AM, pared to a bicarbonate free solution with a Farthing MJG. Efficacy oforal rehydration solutions in a rat comparable glucose concentration (solution IV), model of secretory diarrhea. J Pediatr Gastroenterol Nutr on September 23, 2021 by guest. Protected copyright. 1987; 6: 624-30. confirming the results of our rat intestinal per- 6 Wapnir RA, Lifshitz F. Osmolality and solute concentration - fusion studies.' Bicarbonate loss did occur their relationship with oral hydration solution effectiveness: an experimental assessment. Pediatr Res 1985; 19: 894-8. during whole gut perfusion with bicarbonate free 7 Rolston DDK, Mathew PM, Mathan VI. Food-based solu- solutions, but did not result in clinical acidosis. tions are a viable alternative to glucose-electrolyte solutions for oral hydration in acute diarrhoea - studies in a rat model This is probably because the kidneys are capable of secretory diarrhoea. Trans R Soc Trop Med Hyg 1990; 84: of correcting (within limits) bicarbonate loss 156-9. 8 Feinstein AR. Clinical epidemiology: the architecture of clinical without base precursor supplementation so long research. Philadelphia: Saunders, 1985. as normal renal perfusion is maintained. Indeed, 9 Rolston DDK, Kelly MJ, Borodo MM, Dawson AM, Farthing MJG. Effect of acetate, citrate and bicarbonate on clinical trials which compared bicarbonate free water and sodium movement in normal and cholera toxin- oral rehydration solutions with solutions con- treated rat small intestine. Scand J7 Gastroenterol 1989; 24: 1-8. taining bicarbonate showed that there was no 10 Rolston DDK, Moriarty KJ, Farthing MJG, Kelly MJ, Clark difference in the morbidity, mortality, and purg- ML, Dawson AM. Acetate and citrate stimulate water and sodium absorption in the human jejunum - implications for ing rates in these two groups studied.'629 Exclud- oral rehydration therapy. Digestion 1986; 34: 101-4. ing bicarbonate from oral rehydration salt 11 Rolston DDK, Mathan VI. Xylose transport in the human jejunum. DigDisSci 1989; 34: 553-8. packets would decrease cost and simplify packag- 12 Alvarado F. D-xylose active transport in the hamster small ing, which are important considerations in intestine. Biochim Biophys Acta 1966; 112: 292-306. 13 Powell DW. Intestinal water and electrolyte transport. In: developing countries. Johnson LR, Christensen J, Jackson MJ, Jacobson ED, The effect that flow rate has on water and Walsh JH, eds. Physiology ofthegastrointestinal tract. Volume 2. 2nd ed. New York: Raven Press, 1987: 1267-305. solute transport should be considered when 14 Rolston DDK, Mathan VI. Jejunal and ileal glucose-stimu- interpreting the results of perfusion studies. lated water and sodium absorption in tropical enteropathy: implications for oral rehydration therapy. Digestion (in This is particularly important in the present press). study because although the initial flow rates were 15 WHO/UNICEF Joint Statement. The mnanagemnent ofdiarrhoea and use of oral rehydration therapy. Geneva: World Health identical (30 ml/min), the flow rates of the Organisation, 1985. solutions as they moved down the intestine were 16 Islam MR, Ahmed SM. Oral rehydration solution without bicarbonate. Arch Dis Child 1984; 59: 1072-5. different as evidenced by the different rectal flow 17 Hanig JM, Soergel KH, Barry J, Ramaswamy K. Brush border rates (Table III). Flow rate directly determines membrane vesicles formed from human duodenal biopsies exhibit Na+-dependent concentrative L-leucine and the solute load (initial solute concentration x D-glucose uptake. Biochem Biophys Res Commun 1988; 156: flow rate). This in turn has a pronounced effect Evaluation ofthe efficacy oforal rehydration solutions usinghuman wholegutperfusion 1119 18 Davis RG, Santa Ana CA, Morawski SG, Fordtran JS. 25 Saunders DR, Sillery JK. Absorption of carbohydrate-electro- Development of a lavage solution associated with minimal lyte solutions in rat duodenojejunum. Implications for the water and electrolyte absorption or secretion. Gastro- composition of oral electrolyte solutions in man. Dig Dis Sci enterology 1980; 78: 991-5. 1985; 30: 154-60. Gut: first published as 10.1136/gut.31.10.1115 on 1 October 1990. Downloaded from 19 Dahlquist A. Assay ofintestinal disaccharidases. Anal Biochem 26 Lifshitz F, Wapnir RA. Oral rehydration solutions: Experi- 1%8; 22: 100-3. mental optimization of water and sodium absorption. 20 Hyden S. A turbidometric method for the determination of the J Pediatr 1985; 106: 383-9. higher polyethylene glycols in biological materials. KLantbr- 27 Sladen GE, Dawson AM. Effect of bicarbonate on sodium Hogsk Annl 1955; 22: 139-45. absorption by the human jejunum. Nature 1968; 218: 21 Fordtran JS, Rector FC, Ewton MF. Permeability characteris- 267-8. tics of the human small intestine. J Clin Invest 1965; 44: 28 Turnberg LA, Fordtran JS, Carter NW, Rector FC. 1935-44. Mechanisms of bicarbonate absorption and its relationship 22 Frase LL, Strickland AD, Kachel GW, Kreis GJ. Enhanced to sodium transport in the human jejunum. J Clin Invest glucose absorption in the jejunum of patients with cystic 1970; 49: 548-56. fibrosis. Gastroenterology 1985; 88: 478-84. 29 Clements ML, Levine MM, Cleaves FT, et al. Comparison of 23 Sladen GE, Dawson AM. Interrelationships between the simple sugar/salt versus glucose/electrolyte solutions in absorption of glucose, sodium and water by the normal infant diarrhea. J Trop Med Hyg 1981; 84: 189-94. human jejunum. Clin Sci 1969; 36: 119-32. 30 Sladen GE, Dawson AM. Effects offlow rate on the absorption 24 Binder HJ. Absorption and secretion of water and electrolytes ofglucose in a steady state perfusion system in man. Clin Sci by small and large intestine. In: Sleisenger MH, Fordtran 1969; 36: 133-45. JS, eds. Gastrointestinal disease. Philadelphia; Saunders, 31 Holdsworth CD, Dawson AM. The absorption of mono- 1983: 811-29. saccharides in man. Clin Sci 1964; 27: 371-9. http://gut.bmj.com/ on September 23, 2021 by guest. Protected copyright.