Contents - Previous - Next

This is the old United Nations University website. Visit the new site at


Kenneth H. Brown


Diarrhoeal disease control programmes have expanded remarkably during the past two decades, a period that has also witnessed further progress in relevant scientific research on the pathophysiology and appropriate treatment of enteric infections and their complications. Despite these important advances, diarrhoea remains a major cause of morbidity and mortality in low-income countries. An estimated 1.5 billion episodes of diarrhoea occur annually among children less than five years of age in the developing world, resulting in approximately four million deaths and contributing to secondary malnutrition [1]. Until feasible, low-cost preventive measures can be more widely applied, appropriate case management of diarrhoea, consisting of rehydration therapy, diet, and judicious use of antibiotics, will remain the cornerstone of intervention programmes designed to reduce associated deaths and other complications.

Rehydration therapy with glucose-electrolyte oral rehydration solution (ORS), as promoted by the World Health Organization (WHO), has been an especially valuable component of case management because it is simple, inexpensive, and effective. However, standard forms of glucose-ORS do not reduce the severity of purging or the duration of illness. Moreover, the production, distribution, and use of these solutions still fall far short of estimated needs. Finally, ORS alone does not satisfy energy and nutrient needs, and must be compensated by concurrent attention to adequate dietary therapy [2]. For these reasons, alternative modes of therapy have been sought.

Two general approaches have been taken in an attempt to improve standard glucose-ORS. The first is to incorporate additional organic molecules, such as amino acids and oligopeptides, into standard glucose-electrolyte solution promoted by WHO in order to further enhance sodium transport. In general, this approach to develop a "super-ORS" has not been very rewarding [3], possibly because of the added osmotic burden of these mixtures. The second strategy is to prepare solutions with reduced osmolality by including carbohydrate polymers in place of glucose and other simple sugars and/or by reducing the concentrations of glucose and sodium below those of the WHO formulation. Evaluation of low-osmolality ORS is an active area of clinical research, and several recent studies indicate that these solutions may yield clinically important reductions in faecal output [4].

Preparation of ORS from common food sources of carbohydrate polymers is one of the possible ways to reduce the osmolality of these solutions. Most experience with food-based solutions has been with rice-based ORS, although other cereals, tubers, and legumes have also been used. A recent meta-analysis of 13 clinical trials of rice ORS found that faecal output was reduced by 32% to 36% during the first day of treatment of patients with cholera when rice-ORS was used in place of glucose-ORS [5]. However, stool outputs were reduced by only about half that amount among children with acute non-cholera diarrhoea. Moreover, studies completed since the publication of this meta-analysis found no advantage of rice-ORS for children with acute, non-cholera diarrhoea, possibly because of greater emphasis on dietary therapy during the later studies [6]. Thus, although rice-ORS may be superior to standard therapy for patients with cholera, there may be no clinical advantage of rice-ORS for children with other forms of diarrhoea who receive food along with rehydration therapy.

Despite these variable clinical outcomes, promotion of food-based ORS through diarrhoea control programmes remains attractive for several reasons. Not only do food-based solutions provide the theoretical advantage of reduced osmolality, but also staple foods are readily available at the household level and are often already being used as domiciliary remedies for diarrhoea. By contrast, glucose-containing ORS packets are generally less accessible and may be more costly. Additionally, the energy density of food-based ORS is typically several-fold greater than that of glucose-ORS, and the former also contain small amounts of protein and other nutrients. Thus, food-based ORS has a potential nutritional advantage, although this may be of less importance when a nutritionally adequate diet is offered in addition to glucose-ORS [7].

Based on the aforementioned considerations, many scientists, clinicians, and public health planners have argued that it is time to replace standard glucose-electrolyte solutions with cereal-based ORS [8]. Issues that need to be resolved in particular settings before choosing between the proven benefits of glucose-based ORS and the less certain advantages of the more recently developed cereal-derived products are whether the specific food-based solutions are safe for home use and acceptable to patients and their caregivers. A growing body of scientific studies has addressed these issues, and this issue of the Food and Nutrition Bulletin explores them further in the context of two community-based trials in Guatemala and Mexico. Interestingly, the research teams in each site arrived at somewhat different conclusions regarding the safety and efficacy of their respective rice-based solutions. It is therefore instructive to compare the two sets of experiences to understand which factors may have been responsible for the disparate set of results.

The series of studies in Guatemala focused on the development and application of a rice-ORS in which the concentration of cereal was increased in stepwise fashion in an attempt to enhance the nutritional value of the resulting preparations. The viscosity of the more concentrated mixtures was reduced by adding amylase, thereby assuring that they would be drinkable. The first set of studies in the series examined the appropriate amounts of amylase that would yield the most desirable organoleptic characteristics and osmolality for solutions containing 5%,10%, or 15% (w/v) rice solids [9]. The osmolality of the amylase-treated 10% and 15% solutions was 225 and 295 mOsm/L, respectively, each of which is less than the physiological osmolality of extracellular body fluids.

During the next phase of the project, the investigators assessed the acceptability, tolerance, and efficacy of each of the three rice-containing solutions by measuring the amounts that were consumed during the first four hours of therapy and the initial responses to treatment of young children who had acute diarrhoea and mild or moderate dehydration [10]. The children drank enough of the solutions to supply 70% to 90% of their initial fluid deficits within four hours. The amounts consumed were similar for each preparation, but energy intake was, as expected, considerably greater with the higher-density solutions. Notably, there were three times as many episodes of vomiting and somewhat more instances of therapeutic failures with the most concentrated solution, but the differences were not statistically significant, possibly because of the limited statistical power provided by the fairly small sample size (n = 63). More over, the duration of the protocol was only four hours. It is uncertain whether significant differences would have been identified had the observations lasted longer.

Regrettably, the investigators did not include the results of their longer-duration clinical trial with the current package of studies, although they have previously presented preliminary results of a study in which the three solutions were compared. In that trial they identified an increased rate of therapeutic failure with the 15% solution, but no differences in the risks or benefits of therapy when the remaining two solutions were compared [11].

Following these sets of laboratory and clinical experiences, the investigators conducted additional studies to examine the feasibility of communicating the correct preparation of a complete rice-ORS that could be prepared from home-available ingredients [12]. Salt, orange juice, and baking soda were added to the amylase-treated, 10% rice-ORS to provide sodium, potassium, and bicarbonate. Mothers were interviewed to elicit their opinions on the modified solution, and they were shown how to prepare it, using common household measuring devices. A subset of the mothers who were interviewed at home also received ingredients so they could prepare the solutions themselves.

Nearly all the mothers reported that the solutions were acceptable, but they were frequently unable to prepare them as specified. Several of the home-prepared solutions had dangerously high sodium concentrations, and nearly one-fourth had excessive osmolality. When the modified solution was later tested in an open, uncontrolled trial of 19 mildly or moderately dehydrated children, approximately one-third failed to respond to therapy. Despite the apparent acceptability of the modified solution to children's caregivers, the authors were appropriately skeptical that it could be prepared accurately or that it would be sufficiently efficacious to warrant further promotion in the community.

The approaches and experiences in Mexico were quite different. The investigators began by attempting to identify those oral solutions that were already being used by children's caregivers to treat diarrhoea at home. They reasoned that case management protocols would ultimately be more effective if they relied on existing practices rather than promotion of new therapies. Because effectiveness at the population level is the product of treatment efficacy and adoption of the recommended therapy, they chose to focus initially on strategies to maximize coverage and were willing to compromise, if necessary, on efficacy [13].

The investigators launched their project with a series of anthropological studies designed to describe existing beliefs and practices concerning treatment of diarrhoea [14]. They also collected samples of fluids used to treat diarrhoea at home and analysed their proximate composition, concentrations of sodium and potassium, and osmolality. Caregivers reported that herbal teas and rice-based beverages were both widely used for home treatment; but, unlike the teas, the rice-containing fluids were generally given regardless of the type of diarrhoea and for longer periods of time during illness. The investigators decided to promote the rice beverages for these reasons and also because the caregivers tended to view the teas as remedies to stop diarrhoea, whereas they did not expect the rice solutions to perform this way. Thus, they were presumably less likely to become disillusioned with the therapy if rice-based solutions were adopted. Analysis of the rice solutions indicated that they almost always had osmolalities within an acceptable range, despite frequent addition of sucrose, but they rarely, if ever, contained additional salt. Thus, there was some concern about the possible efficacy of these preparations.

The investigators, therefore, initiated a controlled clinical trial to compare the rice-powder gruel (5% w/v) containing added sucrose (60 g/L) with standard glucose-electrolyte ORS in children hospitalized for acute diarrhoea and mild or moderate dehydration [15]. The rice-based solution contained no additional sources of sodium, potassium, or bicarbonate. Unfortunately, the authors did not specify what diets were given in addition to the ORS, so it is uncertain whether they may have provided some of the aforementioned electrolytes. Interestingly, after patients whose initial faecal purging rates were greater than 10 ml/kg/h were excluded, there were no differences in response to therapy by type of ORS, except that more of the children in the glucose-ORS group had elevated serum sodium concentrations at the time of discharge.

Fortified with the results of the clinical study, the investigators then compared adoption rates of either rice-based or standard glucose-ORS following an intensive, community-based promotional campaign [16]. Similar high rates of use of the respective types of ORS were found in each set of communities. However, caregivers appeared to have slightly greater access to the rice-based solution, which they tended to initiate earlier in the course of the episode. The caregivers were also somewhat more likely to perceive the rice-based solution as beneficial. Although there was a considerable range of concentration of rice and sucrose in the home-prepared solutions, nearly all were within acceptable limits. The investigators concluded from the full set of studies that rice-based mixtures were actually preferable to standard glucose-ORS for home treatment because of the greater availability of the rice preparations.

The experiences with these potentially home-available, rice-based rehydration solutions were quite different in Guatemala and Mexico, with regard to both accuracy of preparation and apparent clinical efficacy. How can these apparently discrepant sets of results be explained? First, it must be appreciated that the two research teams took diverse approaches to developing the final treatment solutions. In Guatemala, the investigators started with a theoretical notion of how to improve standard glucose-ORS by developing ways of enhancing its nutrient content and possible therapeutic efficacy without sacrificing desirable organoleptic properties of the original solution. Only secondarily did they consider the ability to communicate the correct method of preparing the final recipe. By contrast, the Mexican team began by exploring current forms of home-prepared ORS that were already in use in the community. The final Mexican product was much simpler and could be promoted without relying on a detailed recipe, whereas the Guatemalan mixture had six different ingredients (water, orange juice, rice, amylase, salt, and baking soda) and five different units of measurement (soda bottle, bottle cap, soup spoon, teaspoon, and cup). Given this degree of complexity, it is remarkable, in fact, that the Guatemalan mothers were able to prepare the mixtures as well as they did.

It is also important to recognize that the two mixtures developed for home use in the Guatemalan and Mexican studies were different not only from each other but also from the usual forms of rice-ORS studied in most of the previously published clinical trials of cereal-based ORS. The Guatemalan rice-ORS had a higher than usual concentration of rice. Moreover, the orange juice that was added contributed further to the carbohydrate content of the mixture. Together, these two sources of carbohydrates may have exceeded the intestinal absorptive capacity of some children, thus accounting for the limited efficacy that was reported.

The Mexican product also had greater carbohydrate content than usual because it contained sucrose as well as rice, and it further differed from standard ORS because of its lack of electrolyte sources. It is somewhat surprising that this mixture performed as well as standard glucose ORS in the clinical trial, although this may be explained by the possibility that additional sodium was provided by the children's diets in the hospital, thereby compensating for the deficiencies of the rice mixture. If so, it would be important to assure that home diets would be similarly composed and administered. Furthermore, children with especially high rates of purging were excluded from the clinical evaluation protocol. Thus, the results of this study must be accepted with a degree of caution because they may not necessarily apply to children with more severe diarrhoea. Because caregivers would not ordinarily have access to information on their children's faecal purging rates, they would not be able to provide alternative therapy to those with high stool outputs.

In conclusion, the results of these research efforts, together with prior experience reported by WHO [17], suggest that selection of appropriate fluids and foods for preventing dehydration from diarrhoea must rely on considerations of the local acceptability and availability of different preparations as well as their clinical efficacy. In general, any home-available fluid preparations should be suitable for prevention of dehydration in non-dehydrated children, so long as the sodium contents and osmolality of the fluids are not excessive. When dehydration is present, though, a complete solution with an appropriate balance of carbohydrate and electrolytes is recommended. This implies that caregivers must be able to recognize signs of dehydration to be able to decide when a special solution is needed. From a communication perspective, it would be preferable to rely on simple, pre-existing preparations for home therapy of non-dehydrated children than to teach new recipes, especially if they are very complex. Recommendations for fluid therapy should be accompanied by dietary- advice, and indications for seeking medical attention should also be emphasized.

Both sets of investigators should be applauded for their attempts to develop better approaches to treatment of diarrhoea in the home and to make these novel therapies accessible to low-income communities. The combined reports highlight many of the complexities inherent in the development of safe, efficacious, acceptable, and feasible rehydration solutions for home use. The combined anthropological, clinical, and epidemiological research methods that were used to address this set of issues provide an appropriate and laudable model for health-related research in developing countries.



  1. Claeson M, Merson MH. Global progress in the control of diarrheal diseases. Pediatr Infect Dis J 1990; 9:345-55.
  2. Brown KH. Dietary management of acute diarrheal disease: contemporary scientific issues. J Nutr 1994; 124:1455S-60S.
  3. Bhan MK, Mahalanabis D, Fontaine O, Pierce NF. Clinical trials of improved oral rehydration salt formulations: a review. Bull WHO 1994;72:945-55.
  4. International Study Group on Reduced-Osmolarity ORS Solutions. Multicentre evaluation of reduced-osmolarity oral rehydration salts solution. Lancet 1995; 345:282-5.
  5. Gore SM, Fontaine O, Pierce NF. Impact of rice-based oral rehydration solution on stool output and duration of diarrhoea: meta-analysis of 13 clinical trials. Br Med J 1992;304:287-91.
  6. Division of Diarrhoeal and Acute Respiratory Disease Control. Twenty-five years of ORS. Joint WHO/ ICDDR,B consultative meeting on ORS formulation. Dhaka, Bangladesh: ICDDR,B, 1994.
  7. Fayad IM, Hashem M, Duggan C, Refat M, Bakir M, Fontaine O, Santosham M. Comparative efficacy of rice-based and glucose-based oral rehydration salts plus early reintroduction of food. Lancet 1993;342: 772-5.
  8. Elliott K, Attawell K, Wilson R, Hirschhorn N, Greenough WB III, Khin-Maung-U, eds. Cereal based oral rehydration therapy for diarrhoea. Report of the international symposium on cereal based oral rehydration therapy, 12-14 November, 1989. Geneva: Aga Khan Foundation, and Columbia, Md, USA: International Child Health Foundation, 1990.
  9. Vettorazzi C, Solomons NW, Brown KH, Shoemaker C. Amylase-treated rice flour oral rehydration solution with enhanced energy density. I. In vitro studies of viscosity, osmolality, and stability. Food Nutr Bull 1996; 17:98-103.
  10. Vettorazzi C, Mazariegos M, Molina S, de Ramirez I, Solomons NW . Amylase- treated rice flour oral rehydration solution with enhanced energy density. II. In vivo studies of tolerance, energy intake, and rehydration efficacy in the initial treatment of dehydrated diarrhoeic children. Food Nutr Bull 1996;17: 104-9.
  11. Molina S, Vettorazzi C, Mazariegos M, de Ramirez I, Solomons NW. The a-amylase treatment of rice-based oral rehydration solutions allows for higher caloric intake without perceptible benefits in the outcome of acute diarrhea in Guatemalan children. FASEB J 1995;9:A863.
  12. Hudson R, Vettorazzi C, Mazariegos M, Solomons NW. Amylase-treated rice flour oral rehydration solution with enhanced energy density. III. Trials with a home preparation. Food Nutr Bull 1996;17: 110-16.
  13. Martínez H, Habicht J-P. A programme to develop culturally and medically sound home fluid management of children with acute diarrhoea. Food Nutr Bull 1996;17: 117-9.
  14. Martínez H, Meneses LM, Bernard HR, Pelto PJ. Selection of culturally sound home fluid management of infantile diarrhoea in rural Mexico. Food Nutr Bull 1996;17:120-8.
  15. Martínez H, Habicht J-P, Garza C, Mota F. Clinical trial of a rice-powder oral rehydration beverage. Food Nutr Bull 1996;17:129-37.
  16. Martínez H, Bojalil R, Guiscafré H. Promotion of oral rehydration therapy comparing a home-made rice powder gruel and oral rehydration salts. Food Nutr Bull 1996;17:138-44.
  17. Programme for the Control of Diarrhoeal Diseases. The selection of fluids and food for home therapy to prevent dehydration from diarrhoea: guidelines for developing a national policy. Geneva: WHO, 1993.

Contents - Previous - Next