Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Roberta Hudson, Carolina Vettorazzi, Manolo Mazariegos, and Noel W. Solomons
In companion papers in this series, we established the viscosity and osmolality of rice flour solutions in excess of 5% for producing a fluid oral rehydration solution (ORS), and explored the capacity of such an ORS to initiate rehydration in children. Here we assess a home-prepared ORS based on rice flour with augmented energy content (10% rice w/v) and made drinkable by adding a -amylase to reduce viscosity The study was conducted in community and hospital settings in Guatemala. The objectives were threefold: to determine the acceptability of the 10% rice-based ORS (RB-ORS) through interviews with Guatemalan mothers; to determine the feasibility of producing this energy-dense solution in low-income homes; and to assess the efficacy of the RB-ORS in diarrhoeic children. The flavour and colour of the RB-ORS were reported to be acceptable by 96% of 137 mothers; however, 21% said that the consistency was too thick. The osmolality of the samples ranged from 38 to 607 mOsm/L (mean 380 mOsm/L). The solution successfully rehydrated 13 (68.4%) of 19 hospitalized children. We conclude, however, that the consistency of the therapeutic response to this home-prepared modification of energy-dense RB-ORS would be highly variable in practice. In addition, there is a finite and substantial risk of hyperosmolality with the a-amylase treatment, and of inappropriate electrolyte composition with natural ingredients and household measures.
An estimated three and a half million children under age five years die each year of dehydration from diarrhoeal diseases [1]. The mechanism of illness is generally acute loss of fluids and electrolytes, which, in the extreme, leads to death by hypovolemic shock, electrolyte imbalances, or both [2]. rehydration with compensatory fluids is usually a life-saving intervention. Originally, the fluids were administered intravenously, but research on the Indian subcontinent in the late 1960s validated the use of oral fluids for restoring normal hydration due to diarrhoeal dehydration [3-5]. Oral rehydration therapy (ORT) has proved to be a therapeutically successful [6] and cost-effective [7] treatment in preventing deaths caused by diarrhoeal dehydration. It was lauded by Lancet in 1975 as "the most important medical advance of this century" [8].
The World Health Organization standardized an oral rehydration solution (ORS) formula and produced packets of pre-measured dry ingredients, consisting of glucose as the carbohydrate base, sodium chloride, sodium bicarbonate, and potassium chloride. To prevent dehydration, it is important to initiate ORT at the onset of acute diarrhoea, but WHO ORS is not readily available. Despite programmes to distribute it, it was estimated that only 23% of those who need it actually receive it [9]. For this reason, home-prepared solutions that could be made up at the first sign of a diarrhoeal episode have been explored. Glucose (dextrose) and the reagent-quality electrolyte salts in WHO ORS are not household items. Thus, an ORS that can be made with standard ingredients is critical when a solution such as WHO ORS cannot be obtained [10].
A second disadvantage of a pre-packaged glucose-based formula is that the concentration of glucose is limited secondary to hyperosmolality, which can exacerbate diarrhoea, [11]. Thus, the energy content of such a solution cannot exceed 68 to 80 kcal/L. Moreover, since the glucose concentration must remain near 111 mmol/L, the number of sodium cotransporters is also limited.
Many studies support the use of rice as the carbohydrate base in an ORS [12 15]. Rice is found in the majority of households worldwide [13], and rice water has been used for centuries as a treatment for diarrhoea [12]. Not only is rice ubiquitous and inexpensive, but the concentration of a polysaccharide such as rice can be increased without affecting the osmolality of the solution. A higher concentration of the carbohydrate base provides up to three times more calories per litre [12], as well as additional cotransporters for sodium that permit more absorption of water. The additional energy can be critical during repeated episodes of diarrhoea that can affect a child's growth and nutrition status [10].
In the two other articles from Guatemala and the Center for Studies of Sensory Impairment, Aging, and Metabolism (CeSSIAM) in this series, we explored the concept of incubating a -amylase to alter the fluid characteristics of rice flour emulsions. The first paper explored the in vitro rheology and osmolality features [16]. The second took the concept into the hospital setting [17], but in a careful, limited manner to characterize the initiation of rehydration. Here we turn our attention to a formulation that could be made in the home, provided that the enzyme is made available to the household.
This study had three objectives: to test the acceptability of the recipe for a home-produced 10% rice flour among low-income mothers; to determine the fidelity with which the instructions for the recipe could be followed when the ORS was made from basic ingredients in such homes, with special attention to the rate of hyperosmolar or electrolyte-imbalanced preparations; and to conduct a phase I trial to assess the ability of the rice-based ORS (RB-ORS) prepared from home-available ingredients to aid in the recovery of children with mild to moderate dehydration in a hospital outpatient setting.
The 10% rice-based formula was adapted from the work reported in this series [16,17]. The solution included Clarase L-40,000 a -amylase (Solvay Enzymes, Inc., Elkhart, Indiana, USA) and common household ingredients consisting of rice flour, oranges, baking soda, and table salt (table 1). The initial amounts totaled 1,070 ml. but the final volume was approximately 1,000 ml after evaporation. With the exception of the amylase enzyme, all ingredients were available year around and accessible in local markets. The rice flour was a Guatemalan product, and the local salt was more coarse and moist than that commonly found in the United States.
TABLE 1. Formula of rice-based oral rehydration solution
| Ingredient | Amount |
| Water | 760 ml |
| Orange juice | 300 ml |
| Rice flour | 100 g |
| Clarase 40,000 a -amylase solutiona | 10 ml |
| Baking soda | 2.5 g |
| Salt | 1.75 g |
a. 1.5 ml a -amylase in 12.5 ml distilled water.
The questions on the instrument related to age, parity, frequency of diarrhoea in children, current treatment, acceptability of the flavour, colour, and consistency, and perception of acceptability by a child. A preliminary questionnaire was pretested with 23 subjects and revised to enhance clarity. The final instrument was administered to 82 mothers or caretakers in the waiting room of San Juan de Dios Hospital located in downtown Guatemala City, and to 55 mothers or caretakers in the small periurban community of La Peronia.
The mothers or caretakers ranged in age from 16 to 64 years, and they had a median of three children. They were given samples to taste and then orally responded to the questionnaire that was read aloud in Spanish by a native Guatemalan. To determine the precision of the instrument, 41 of the 55 women in La Peronia were visited and administered the same questionnaire again, in the same manner and by the same interviewer. Care was taken to visit the mothers in the morning hours on both occasions.
After the questionnaire was completed, the author demonstrated the method of preparing the RB-ORS. The method was demonstrated on three separate occasions to approximately 65 mothers. Because no scales or other precise measuring utensils were available in this indigent population, compromises were necessary to measure the ingredients appropriately. The water was measured with empty 10-ounce soda bottles (typically Pepsi-Cola) because the beverage was popular in the settlement and sold at every store. The baking soda and salt were measured with a cap from a soda bottle. In the laboratory, these improvised measuring implements produced an average weight of 2.54 g baking soda and 1.84 g salt.
Flour was measured with household spoons. The instructions emphasized using a large (soup) spoon as opposed to a teaspoon. When measured in the laboratory, the rice flour weighed an average of 19.6 g, which was 98% of the desirable weight of 20.0 g. The a -amylase was premeasured, but the mothers were instructed to use two household teaspoons to measure it. Finally, orange juice was measured with a drinking cup, such as a typical coffee cup.
Forty-four (68%) of the 65 mothers in La Peronia volunteered to participate and were given the necessary ingredients to prepare the solution in their homes. On the same day of preparation, we visited the mothers and collected the samples of the formula. At that time the mothers responded orally to a short questionnaire regarding the steps they took to prepare the solution. Each aliquot (two samples per mother) was labeled at the moment of collection and subsequently centrifuged at San Juan de Dios Hospital. After centrifugation, the samples were placed in a freezer until transported in an ice chest to a laboratory at the general hospital of the Guatemalan Institute of Social Security for analysis by an osmometer for potassium and sodium concentrations and osmolality.
In the final phase of this study, the solution was given to 20 boys who came to the emergency room of San Juan de Dios Hospital with a dehydrating diarrhoeal episode. Administration followed the general approach and procedures of a CeSSIAM project that had been conducted in the same hospital facilities [18]. The mean age of the children was 16.4 months (range 3-36 months). The frequency of vomiting and diarrhoea before admission to the hospital, the child's weight, and the percentage dehydration were recorded. The degree of dehydration was determined by a physician, and the appropriate amount of solution was given orally by a nurse or physician. Throughout the four-hour study, the following were observed and recorded for each child: frequency of diarrhoea and appearance of the stools, frequency of vomiting, number of times that the rehydration solution was offered, and volume of solution that the child consumed. At the end of the four hours, the child's final weight and percentage of initial degree of dehydration were determined.
Acceptability of the home-prepared RB-ORS
After a sample of the RB-ORS had been tasted by 137 women in La Peronia, the subjects were asked if they liked or disliked its flavour and colour (table 2). Responses of "yes, a little," "yes," and "yes, very much" were grouped together as acceptable, and the response of "no" as not acceptable. Nearly all mothers found the solution acceptable in colour and flavour, and most rated the consistency as acceptable. In the test-retest series, the percentages for concordance of answers concerning flavour, colour, and consistency were 78%, 88%, and 95%, respectively.
Only three mothers (2%) who were interviewed stated that they would not give the solution to their children as treatment for diarrhoea, and six (4%) did
TABLE 2. Percentage of positive and negative responses of mothers when asked if colour, flavour, and consistency of the solution were acceptable not think that their children would drink the solution. The majority of subjects had no suggestions for changing the solution, but 17% offered specific preferences for more flavour and thinner consistency.
| Organoleptic property | Yes | No |
| Colour | 99 | 1 |
| Flavour | 96 | 4 |
| Consistency | 77 | 23 |
Consistency and accuracy of the preparation of the RB-ORS
Table 3 shows the sodium, potassium, and osmolality values of the rice flour solutions prepared by the mothers. Of the 44 samples, 3 (7%) had sodium concentrations that were considered to be dangerously high [19], namely, 120 mmol/L or greater. The sodium contents of these samples were 124,138, and 186 mmol/L. If we consider the desirable range to be 40 to 60 mmol/L, 3 samples were below the recommended level and 27 were above it, with 14 falling within the limits. Each sample had a distribution of 4 specimens with potassium concentrations above 25 mmol/L and 11 below 15 mmol/L. No sample had a potassium concentration of 35 mmol/L or greater, which some consider dangerously high [20]. Ten of 41 samples measured had an osmolality in excess of 440 mOsm/L, which some believe to be in the dangerous region [21]. Another 22 exceeded 300 mOsm/L, and only 23% were in the desirable range of 170 to 300 mOsm/L.
The sodium and potassium concentrations and osmolalities did not significantly differ between the solutions prepared by the 32 women who received both oral and written instructions and those prepared by the 12 women who received only oral instructions. As indicated in the postpreparation questionnaire, of the 16 mothers who responded that they used the correct procedures, the samples of only 6 were within the acceptable range for sodium, potassium, and osmolality.
Ability of RB-ORS to rehydrate hospitalized patients
Of the 20 boys treated in the hospital, all but 1 were diagnosed with mild or moderate dehydration; 7 had up to 5% loss of initial body weight and 12 had a 5.0% to 9.9% deficit. One subject was excluded from the analysis due to a false diagnosis of dehydration.
TABLE 3. Electrolyte and osmolality values of rice-based oral rehydration solution (RB-ORS) prepared by mothers in Peronia
| Component | Goal for 10% RB-ORS | WHO ORSa | N | Mean | Median | Range | Dangerously high samplesb |
| Sodium (mmol/ L) | 40-60 | 90 | 44 | 71.1 | 69.5 | 19.1-186.1 | 3 |
| Potassium (mmol/L) | 15 25 | 20 | 44 | 17.5 | 17.6 | 7.7-30.8 | 0 |
| Osmolality (mOsm/L) | 170-300 | 331 | 41 | 377 | 381 | 183-759 | 10 |
120
mmol/L [19]; potassium 35 mmol/L [20]; osmolality 2440 mOsm/L [21].Thirteen subjects were successfully rehydrated with the RB-ORS. Six were declared treatment failures and shifted to an alternative rehydration therapy.
The frequency of vomiting before the study and the age were similar in those who were rehydrated with the RB-ORS and those who were not. However, the number of faecal evacuations in the six children who were not rehydrated was significantly greater before admission than in those who were rehydrated. Of the six who were considered treatment failures, one child vomited a total of 15 times during the four-hour study. The remaining five experienced neither severe diarrhoea, (they had 2 or fewer evacuations) nor excessive vomiting throughout the treatment period.
Acceptability
Of the three organoleptic variables, only the texture of the liquid caused any degree of concern. The WHO ORS glucose-electrolyte solution has the consistency of water and is widely accepted by mothers. Our liquid is even thicker than other common cereal gruels that the mothers may prepare, as the amylase action allows more carbohydrate to go into suspension. Mothers may equate diluteness with efficacy in treating diarrhoea, as heavy foods are traditionally avoided during diarrhoeal episodes. Nevertheless, all mothers said that they would use the solution as therapy for diarrhoea, and all but six believed that their children would accept it.
Home preparation of RB-ORS
Home preparation of ORS was of limited success, as evidenced by the analysed characteristics of osmolality and electrolyte composition. One explanation for this may be the improvised measuring implements. Makeshift utensils were chosen so that all mothers would have access to them. But it appears that the precision of the utensils or the mothers' ability to use them accurately varied considerably. Other studies also showed that different volumes (or lack) of household utensils contributed to the variability of electrolyte concentrations [20-23]. Although we succeeded in obtaining desirable ranges of concentrations with the soda bottle cap in the laboratory, the mothers in La Peronia were not as successful, as seen in the range of sodium concentrations and the percentage of samples with dangerously high sodium levels.
Since fewer than one-third of the mothers stated that they followed the correct procedures, behavioural causes for unsatisfactory solutions are more reasonable. Since the mothers received verbal instructions, demonstrations, and, in some cases, written instructions, the low percentage of those who were able to follow the directions correctly was troubling. It may be that their education level did not permit consistency with a formulation as complicated as the RB-ORS, or cultural resistance to the recommended proportions may have emerged when mothers were required to prepare the solution on their own.
The timing and heat necessary to inactivate the a-amylase are crucial to obtaining the desired osmolality. If undercooked, the amylase enzyme could continue to hydrolyse the polysaccharides and cause the osmolality to increase due to the continuing release of glucose molecules. For example, two samples measured had osmolalities of 759 and 228 mOsm/L, compared with an ideal below 330 mOsm/L. As the sodium and potassium concentrations were 125 and 26 mmol/L and 110 and 20 mmol/L in the two samples, respectively, we can only attribute this wide divergence to variance in the carbohydrate component, namely, degrees of glucose liberation from rice starch. In preparing the former sample, the enzyme may not have been inactivated because of inadequate heating time or temperature. The absence of a clock or watch in the household may explain any inadequate cooking time.
Salt for the ORS was measured with a soda bottle cap, and sodium concentrations ranged from 19.1 to 186 mmol/L. In one study, 205 mothers in a village of Bangladesh were instructed to use a teaspoon to measure the salt per litre of water, and the resulting sodium concentrations ranged from 2 to 250 mmol/L [24]. Although that sample was approximately five times larger than ours, the range of sodium concentration is similar. In 1981 a programme with a very strong educational component was created to teach women in a Bangladeshi village how to prepare a sugar-salt rehydration solution [25]. Samples with sodium concentrations above 120 mmol/L were considered dangerous. Only 4 (0.4%) of 1,016 samples were at that danger level, compared with 7% of 44 samples in our study. This underlines the importance of educating the mothers adequately and catering to the education level of the specific village [25]. Having the personnel in place to provide such intensive education, however, would almost guarantee infrastructure to supply the WHO ORS packets to the households of the catchment area.
Moreover, previous studies indicated that hypotonic formulas may cause greater absorption of fluid than isotonic formulas [26-28]. Since two-thirds of the homemade formulas were above 330 mOsm/L, these rice flour solutions prepared by the women of La Peronia were not hypotonic.
Clinical trial with the 10% RB-ORS
With the community emphasis of the first two components, and the rationale for this RB-ORS formulation for preparation in the home, it would have been logical to test the efficacy of the solution in a community context in homes or health posts. However, for logistical and ethical reasons we used a well-managed hospital setting for a phase I trial. The location should concentrate a relatively uniform variety of diarrhoea cases with dehydration, and a public referral hospital enhances the efficiency of enrollments. Also, the medical teams in a hospital provide careful measurements of change in hydration status and intensive skilled clinical evaluation [18]. Preparation of the formula by hospital professionals, moreover, ensured that the desired levels of electrolytes and osmolality [17] were achieved even using household ingredients. Finally, given the uncertainties of the efficacy of this novel rice-based solution in therapeutic rehydration, the safety of subjects was better ensured in the context of a fully equipped medical facility.
The results of the clinical trial suggest that a 10% rice-based formula with exogenous amylase can have only limited success in treating mild to moderate dehydration, since 13 of 19 participants were rehydrated with the RB-ORS, for a failure rate of 31.6%. This therapeutic success over the duration of a full course of treatment was less than that seen in the initial phases of treatment for concentrations up to 15% [17]. That companion study used the same principles but an RB-ORS made from reagent-grade ingredients rather than common household items. Since this was a phase I study, we had no simultaneous control group receiving a standard therapy.
In a historical context, however, 17 (89%) of 19 adults with cholera were rehydrated with the rudimentary glucose-based ORS [3]. In an early trial of sucrose-based ORS in Calcutta [29], 19 (95%) of 20 children with moderate to severe diarrhoeal dehydration (no controls) were rehydrated successfully. Thus, for a free-standing, uncontrolled trial, our recovery rate was relatively modest.
It is a matter of speculation why our rates of rehydration success did not equal those of earlier phase I trials. Neither excess vomiting nor excess purging explains the failures. Possible explanations are the limited ability to drink rapidly because of the solution's density, or the sense of satiety that masks or counteracts the thirst driven by the dehydration.
Judgements about the safety and efficacy of alternative approaches to rehydrating children with diarrhoeal dehydration must ultimately be made in a pragmatic and programmatic context. Oral rehydration is a life-saving technique for a disease that, when untreated or inappropriately treated, could have fatality rates of up to 10%. In the largest context, in countries characterized by low education rates and lack of resources, any effort to replace lost and purging fluids with oral beverages is intrinsically better than nothing.
This was the operative dichotomy at the beginning of the oral-rehydration era. Similarly, almost any fluid-replacement strategy is preferable to any use of antidiarrhoeal drugs aimed strictly at the purging [30]. In the starkest relief, even if 2% of subjects might die of an episode as a result of an intrinsic danger from osmolality or electrolyte imbalance of an oral beverage itself, if 80% of the children so treated were saved from an otherwise fatal outcome, the net public health gain would be to reduce fatality from 10% to 2%. This is not insubstantial, but the medical community always strives to improve, even on a good thing.
That oral rehydration using the standard WHO formula is the major medical advance of the century [8] may be unquestioned, but analysts have identified ways its impact can be improved. Moreover, that search for improvement has a complex agenda. Many points in that agenda were addressed in these studies. First, the poor accessibility and availability of prepackaged glucose and salt mixtures [9] led to the search for alternatives that would allow ORS production from household ingredients [31]; this search has moved from sucrose to rice. Second, the putative benefits of rice in reducing purging [32] are of interest, and one must test the hypothesis that the more rice the better. The final concern, beyond increasing survival and reducing the course of the disease, relates to nutrition. Each episode of diarrhoea represents a catabolic stress, which when combined with the caloric restriction that may occur from traditional 2% glucose ORS, can accumulate over several episodes to result in permanent growth failure. Aggressive energy replacement throughout the course of each illness has been proposed to have beneficial effects on growth and general nutrition [10]. We sought to look to the ORS itself, not just to complementary feeding, to enhance caloric intake in the intraepisode interval.
If we analyze our results in terms of the elements of this agenda, the conclusions from the CeSSIAM series offer some stark reservations. Making precisely dosed a-amylase available at the household level has the same logistical problems as providing WHO ORS packets. Even if all of the other ingredients are at hand in the home or local shop, one must have a precisely measured exogenous supply of the enzyme. In Bangladesh [33] and India [34], attempts to liquefy energy-dense, rice-based weaning foods used germinated wheat rather than a commercial chemical as the amylase source. Even without the cooking time required to deactivate the enzyme, getting a -amylase to the household and the vagaries of the shelf-life of the enzyme would cause problems.
Moreover, making the recipe correctly, even with access to the enzyme, seems to be a limitation. With low levels of literacy, one would expect improved reproducibility of recipe preparation with repetition. Since diarrhoea may be a relatively rare event in given households, these mothers would not make up the solution often. In our study, the mothers were asked to prepare the formula soon after they received their instructions, and the results were poor. We can imagine that with increasing time, the accuracy with which they followed the instructions would diminish even more.
Partial hydrolysis of rice starch allows one to get more rice into a given volume for whatever antipurging and nutritional benefits might be gained. But the potential for dangerously high osmolality emerges when the preparation leaves the careful control of the hospital setting [17] for the community households themselves.
Thus, despite its high acceptability and the statement by a majority of mothers that they would use this RB-ORS if it were available, mothers cannot be counted on to make an accurate and consistent solution in their own homes, even when pre-measured enzyme is distributed. The 68.4% of children who were rehydrated with the formula in hospital is low compared with historical experience with other varieties of glucose- or sucrose-based ORS. Given the variability in the use of other ingredients, relying on in-home measurement of the enzyme, the most critical component, would not be prudent. However, providing pre-measured aliquots would involve the same difficulties as the method to be replaced, that is, the WHO ORS packets. Research in the field of oral rehydration should attempt to complement, rather than replace, existing formulas. Approaches to minimize the nutritional impact of diarrhoea, such as continued feeding, must also be adequately explored and promoted.
The authors are grateful to Dr. Susana Molina of the San Juan de Dios general hospital and to the social services and nursing staffs that collaborated in a number of the components. Dr. Hector Caballero is acknowledged for his clinical participation in the rehydration trial in the outpatient Rehydration Unit of that hospital. We thank Dr. Pablo Yurrita of the IGSS hospital for providing access to and assistance with the electrolyte measurements. Lic. Fabiola Micheo and Lic. Carmencita Arriola from the Central American Institute of Investigations and Industrial Technology in Guatemala provided instruction about and access to the osmometer. Prof. Jeanne Freeland-Graves, head of the Department of Human Ecology at the University of Texas at Austin, is appreciated for her institutional support of the project.