Contents - Previous - Next

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

Lower prevalence of diarrhoea in young children fed lactic acid-fermented cereal gruels

Wilbald Lorri and Ulf Svanberg



In many developing countries, diarrhoea is one of the major precipitating factors of child morbidity and mortality. It also significantly contributes to the high prevalence of malnutrition in young children. Available information suggests that a major part of all diarrhoea episodes are associated with bacterial contamination of weaning foods. In vitro studies, however, have shown a strong inhibitory effect of lactic acid-fermented weaning gruels on the proliferation of diarrhoea-causing pathogens.

To assess whether this inhibitory effect is reflected by a lower transmission of diarrhoea, a prospective epidemiological study was carried out to measure the prevalence of diarrhoea in a group of young children in Tanzania who were given a lactic acid-fermented cereal gruel on a daily basis.

Two groups of about 100 children under five years of age were selected on the basis of the use or non-use of fermented gruels. The nutrition status of the children in each group, as measured by comparison of their weight for age with the standard, was not statistically different (p > .05). The mean age of the children in the two groups was similar. An average feeding frequency of 3.5 times per day was reported for both groups. Biweekly recall for diarrhoea, defined as watery, loose stools two or more times per day, was conducted for nine months among selected children.

The mean number of diarrhoea episodes over the study period was 2.1 for children eating fermented gruels, compared with 3.5 for those eating non-fermented gruels (p<.001). The frequency of diarrhoea was also age dependent: it was significantly higher in children under three years of age than in older children in both diet groups (p<.001). Fever episodes were reported in similar proportions in both groups, and were significantly related to diarrhoea (p < .001). Other diseases and the children's nutrition status were not related to the frequency of diarrhoea.


Diarrhoea is one of the major factors that contribute significantly to high child morbidity and mortality in many developing countries [1]. Among the causative agents, the following bacteria have been reported: enterotoxigenic Escherichia coli (ETEC), Shigella, Salmonella, and Campylobacter [2-4]. Among the viruses, rotavirus seems to be the most common [5].

Food contamination is one major route for the transmission of enteropathogens, especially under the hygienic conditions prevailing in a rural setting. Various studies have reported that the source of enteropathogens was either water or food [2, 5-11]. For most people in developing countries, the major source of food is cereals, and dairy products are limited to a very small segment of affluent groups. Presumably, the reports of food as the origin of diarrhoea refer to cereal-based diets, since all the cases cited came from developing countries.

Lactic acid fermentation, a traditional household-level technique, reportedly is effective in reducing or eliminating the growth of diarrhoea-causing pathogens. Possible antagonistic effects of lactic acid-producing bacteria on the pathogens that have been proposed include organic acid production, competition for nutrients, hydroperoxide formation, bacteriocins of protein nature, and antibiotic-like substances [12]. The potential control of intestinal pathogen growth by lactic acid bacteria has received great attention. Much of the reported work has concentrated on Lactobacillus acidaphilus [13-18], which is both preventive and therapeutic in controlling intestinal infections. Other studies have shown the potential health benefits of other species of lactic acid-producing bacteria [19-23]. Most of the studies evaluating the antagonistic effect of lactic acid producing bacteria on intestinal pathogens have been based on dairy products. Although information on the effect of these bacteria from cereal-based products on the control of diarrhoea in the community is very limited, a few in vitro studies have reported inhibited growth of ETEC and diarrhoea-causing pathogens of genera such as Campylobacter, Salmonella, Shigella, Staphylococcus, and Bacillus [24-28].

The aim of this study was to assess whether or not the inhibitory effects of enteropathogenic bacterial growth exhibited under laboratory conditions are also reflected by lower frequency of diarrhoea among young children in a village setting who eat fermented gruels as part of their daily diet.



A baseline survey for the assessment of children who eat fermented gruels was carried out in two adjacent villages, Ilunda and Mudida, in the Iramba and Singida districts respectively, Singida region, Tanzania. The information collected from each family included the weight and age of the children, the types of food given to the children, including the use or not of fermented gruels, feeding frequency per day, and the source of water for domestic use. The general hygienic conditions were also observed.

The two villages are located about four kilometres apart. Ilunda is inhabited mainly by the Wanyiramba tribe, traditional users of fermented gruels (magai) on a daily basis, whereas in Mudida the inhabitants, of the Wanyaturu tribe, use no fermented gruels at all. The reasons for using or not using fermented gruels stem from tribal backgrounds and traditions. A majority of the Wanyaturus are Muslim, and they associate fermented (although non-alcoholic) gruel with the brewing of beer. Some consider it "old-fashioned" food. The Wanyiramba, on the other hand, have used magai throughout the history of their tribe. The majority are Christian, and they do not associate fermented gruel with alcoholic fermentation. All age groups use it as a thirst quencher throughout the year, as the area is generally semiarid.

Although their dietary habits differ, the two tribes live in villages located in a similar agro-ecological zone. They obtain water from similar sources, small streams and pools. The hygienic conditions of the villages are more or less similar. Both villages usually have a surplus of food, as the climate is favourable for growing maize and sorghum.



A sample of 105 children under five years of age was randomly selected in each village with the assistance of village leaders called "ten-cell leaders" as they are responsible for a group of ten households in the immediate area. Each leader was asked to prepare a list of children under five years of age in his or her group, and every fourth child from the list was chosen (the range was 20-40 children for each group of ten households). In Ilunda, where lactic acid-fermented magai is used, 102 of the children completed the study; in Mudida, where non-fermented products are used, 99 children completed the study. (Three and six children respectively dropped out, mainly because their parents either moved to another village or travelled to distant relatives for more than two months and, on their return, they were not re-included in the study.) The children in the two groups who completed the study had a similar age distribution pattern (table 1).

TABLE 1. Age distribution of children in the two diet groups

Age (months) Number of children
Non-fermented diet group Fermented diet group
0-12 13 12
13-24 28 30
25-36 30 23
36-48 22 23
>=49 6 14
Total 99 102

After the children were selected, the parents were convened in a meeting. It was explained that the aim of the study was to examine the children's growth over time in relation to their feeding frequency and food intake. Prevalent diseases afflicting the children from time to time were also to be recorded during the study period and related to their growth and health. Diarrhoea was not emphasized in order to avoid biased answers on the part of the parents during the study. Furthermore, the parents were advised to report to a nearby hospital (Iambi) or dispensary (Mudida) in case the children contracted a disease; no intervention measures were provided by the study team.


The study

The follow-up study began in January (rainy season) and concluded in September (dry season) 1990. The community health workers based in each of the two villages made house calls to the selected households during the second and fourth week of each month, accompanied by one of the authors once a month. The information collected during the baseline survey and the follow-up period included the name, age, and sex of the child: the presence of diarrhoea over the preceding two weeks, defined as watery, loose stools twice or more per day; the use or non-use of fermented foods; the presence of fever over the preceding two weeks; other diseases; the number of meals per day; and the weight of the child (end of the month).

The children were weighed using Salter and bathroom scales. As a nutritional indicator, a standard deviation (SD) score of weight for age was calculated for each child [29], using the following formula:

SD score = (Wsubj - Wref)/SDref

where Wsubj is the weight of the subject, Wref is the median weight for age of the reference population, and SDref is the standard deviation of the reference value (above or below the median as appropriate).


Preparation of the gruel

Magai, a thin porridge or gruel of 10% flour from sorghum grain, is prepared from a mixture of non-germinated and germinated flours (4:1 or 2:1). The germinated flour is obtained by soaking the grain overnight, pouring off the water, and keeping the seeds to sprout in a container covered with wet cotton cloth. The sprouted seeds are spread on the ground to sun-dry and then pounded to a flour using a mortar and pestle. After the porridge is cooked and cooled, about 5% of germinated sorghum flour (total dry matter) is sprinkled over it and stirred in. The mixture is left to ferment overnight or 12 hours. The pH of the product is between 3.8 and 4.0.

The fermented product is used as a drink for one to two days. Children drink it as often as it is available; mother-reported frequency is no fewer than five times a day.


Statistical analysis

The results were evaluated using a multiple linear regression model [30], with the number of diarrhoea cases as a dependent variable. The independent variables were diet (fermented/non-fermented), number of fever episodes, age in months, number of other diseases, and nutrition status (SD score).

Ndiar = X + ▀1 diet + ▀2 fever + ,▀3 age + ▀4 diseases + ▀5 SD score


Ndiar=number of diarrhoea episodes reported over the study period,
X= constant,
diet = use of fermented magai (-1) or not (+1),
fever = number of fever episodes,
age = age of the child in months,
diseases = number of other diseases reported,
SD score = nutrition status at the beginning of the study, expressed as a standard deviation value of weight for age.

The chi-square test was used to evaluate the relationship between two factors.


Children from the two diet groups were matched for age. The weight distribution for each group, expressed as the SD score, is shown in figure 1 (see FIG. 1. Distribution of children by weight for age, expressed as standard deviation (SD) score, in the non-fermented (white; N=99) and fermented (shaded; N=102) diet groups). The number of children with an SD score above -0.8 was slightly higher in the fermented diet group. The overall mean SD score value at the beginning of the study was -0.75 0.72 in the fermented diet group, not significantly different from the -0.95 +0.80 in the non-fermented group (p > .05). Using a multiple regression model, it was found, however, that nutrition status was not correlated to the frequency of diarrhoea (table 2). The results from the multiple regression analysis also indicated that the type of food (non-fermented or fermented), fever, and age were significantly associated with diarrhoea (p < .001), whereas the occurrence of other diseases was not.

TABLE 2. Multiple regression analysis of the number of diarrhoea episodes over the ninemonth study period and various independent factors

Independent variable Value t value Probability
Constant 3.43    
Diet (fermented/non-fermented) 0.54 5.4 .000
No. of fever episodes 0.21 3.3 .001
Age (months) -0.04 6.0 .000
No. of other diseases -0.05 0.8 .4
Nutritional status (SD score) -0.03 0.2 .8

Standard error of estimate =1.33. R2=35 N=201.

The average number of diarrhoea episodes reported was 2.1 in the fermented diet group, which was significantly lower than the 3.5 in the non-fermented diet group (p< .001) (see FIG. 2. Number of diarrhoea episodes in children in the two diet groups over the nine-month study period). About 16% of the children in the fermented diet group reported no diarrhoea episodes, whereas all the children in the non-fermented diet group reported diarrhoea once or more during the study period. Table 3 shows the number of diarrhoea episodes in the different age groups, using aggregated data for both groups. The frequency of diarrhoea declined steadily with increased age (see FIG. 3. Number of diarrhoea episodes per child over the study period in relation to age (y = - 0.044x + 4.16; r2 = .41; p < .001)), from about 4 in children under one year old to about 2 in children over four years old (p < .001).

TABLE 3. Episodes of diarrhoea affecting children in different age groups for both diet groups

Diarhoea episodes No. of children by age (months)
0-12 13 24 25-36 37 48 >=49
0-1 1 7 10 16 9
2-3 9 32 24 24 10
4-5 11 13 15 5 1
>=6 4 6 4 0 0
Total 25 58 53 45 20

X2=36.8. df=112. p=.000.


These findings support the results from in vitro studies [25, 26, 28] that reported a strongly inhibited growth of diarrhoea-causing pathogens with the use of lactic acid-fermented cereal-based foods. In Ghana, fermentation of cereal-based weaning foods reduced gram-negative bacterial contamination from 106 to 104 per gram of gruel compared with non-fermented foods [27]. There are no studies reported, however, on the role of such foods in controlling diarrhoea. In contrast, several studies have reported on the role of dairy products, especially milk products, for the control of diarrhoea [13-15, 18]. Similarly, pure strains of lactic acid-producing bacteria have been isolated from dairy products and used to treat diarrhoea [31, 32]. Other similar studies [10, 18, 33, 34] have reported on the successful use of lactic acid-producing bacteria as a treatment agent, although no change was seen in the duration of diarrhoea in young children treated with a mixture of Lactobacillus and Streptococcus [35].

The number of diarrhoea cases was higher during the rainy and maize-roasting seasons starting in February in both diet groups (see FIG. 4. Prevalence of diarrhoea in the two diet groups during the two-week period preceding each follow-up visit). However, the fermented diet group had a lower percentage of reported cases than the non-fermented group (p<.001). The high peak during the rainy season could be explained by the fact that the water obtained from water pools and valleys during this time is contaminated by a surplus of unclean surface water. The peak in the maize-roasting season is probably due to roasted maize cobs being handled with unclean hands and being directly exposed to the ground. The lower prevalence of diarrhoea in the fermented diet group may be explained by the frequent use of magai by these children. Even if the lactic acid-fermented gruel is contaminated with diarrhoea-causing pathogens, their further growth is likely to be inhibited [25-28] and will not reach levels causing food-borne infections. The frequent use of magai probably also results in less frequent use of non-fermented gruels or snack foods, which are likely to transmit infectious doses of different enteropathogens.

Children under three years of age had more diarrhoea episodes during the study period than older children, with an average number of episodes of 2.1 and 3.5 in the two diet groups respectively. This finding is comparable with that reported by Chen [10], who found diarrhoea frequency peaking between the ages of six months and about three years and who also reported an average of four episodes per year.

The number of fever episodes reported was highly correlated with diarrhoea (p<.001), each one contributing to a 20% increase in the average number of diarrhoea episodes in the regression model. Although the total number of fever episodes was higher in the nonfermented diet group, 358 versus 261, fever reported simultaneously with diarrhoea was about 35% in each group. The association of the two seems likely, as diarrhoea-causing pathogens may also cause fever. Figure 5 (see FIG. 5. Prevalence of fever in the two diet groups during the two-week period preceding each follow-up visit) shows the prevalence of fever during the study period in both groups, and the pattern is similar to that of the diarrhoea prevalence.


The prevalence of diarrhoea observed in this study in young children living under village conditions in Tanzania is comparable to that found in other studies in developing countries, which have reported about four episodes per year in preschool children. A group of young children living in one village in Tanzania who consumed lactic acid-fermented cereal gruels once or more a day had a 40% lower frequency of diarrhoea over a nine-month period than young children living in a nearby village not using fermented gruels.

Diarrhoea was season-dependent in both groups of children, with the highest prevalence during the rainy season and the lowest during the dry season.

It seems likely that fermented gruels do not transmit diarrhoea pathogens, and that a frequent use of this type of food means less frequent use of other food items that are more likely to transmit enteropathogens. It cannot be determined from this study whether the consumption of lactic acid-fermented foods has a therapeutic or prophylactic effect that could explain the lower frequency of diarrhoea in the children eating fermented gruels. It is therefore recommended that a controlled village study should be conducted to evaluate the mode of diarrhoea-lowering action of the frequent consumption of fermented gruels, as well as the aetiological agents of diarrhoea-causing pathogens in young children.


We express our sincere thanks to all the mothers from the study villages, who kindly and untiringly cooperated with us during the nine months of the follow-up period, and to Sven-G÷ran Eriksson of SIK, the Swedish Food Research Institute, for his kind assistance in the statistical analysis of the data.

This work was supported by grants from UNICEF-Tanzania, the International Foundation of Science (IFS), the International Program in the Chemical Sciences (IPICS), Sweden, and the Swedish Agency for Research Cooperation with Developing Countries (SAREC).


  1. UNICEF. The state of the world's children 1988. Oxford; UK: Oxford University Press, 1988.
  2. Black RE, Huq I, Merson MH, Alim A, Yunus MD. Incidence and severity of rotavirus and E. coli diarrhoea in rural Bangladesh. Lancet 1981;1:141-43.
  3. Gross RJ. E. coli diarrhoea. J Infect Dis 1983;7:177-93.
  4. Suan GJ, Rowland GDH, Lloyd-Evans N. Williams K, Rowland, MM. The etiology of diarrhoea studied in the community in young urban Gambian children. J Diarrhoeal Dis Res 1985;3(1):7-13.
  5. Black RE, Merson MH, Rahman ASMM et al. A two year study of bacterial, viral and parasitic agents associated with diarrhoea in rural Bangladesh. J Infect Dis 1980;142:66064.
  6. Rowland MOM, Barrel RAE, Whitehead RG. Bacterial contamination in traditional Gambian weaning foods. Lancet 1978;1:136-38.
  7. Jiwa SFH, Krovacek K, Wadstr÷m T. Enterotoxigenic bacteria in food and water from an Ethiopian community. Appl Environ Microbiol 1981;41(4):1010-19.
  8. Agbonlahor CE, Odugbemi TO. Enteropathogenic, enterotoxigenic and enteroinvasive E. coli isolated from acute gastroenteritis patients in Lagos, Nigeria. Trans Roy Soc Trop Med Hyg 1982;76:265-67.
  9. Black RE, Brown KH, Becker S. Abdul Alim ARM, Merson MH. Contamination of weaning foods and transmission of enterotoxigenic E. coli diarrhoea in children in rural Bangladesh. Trans Roy Soc Trop Med Hyg 1982;76:259-64.
  10. Chen LC. Interactions of diarrhea and malnutrition: mechanisms and interventions. In: Chen LC, Scrimshaw NS, eds. Diarrhea and malnutrition. New York: Plenum Press, 1983:319.
  11. Mathur R. Reddy V. Bacterial contamination of infant foods. Ind J Med Res 1983;77:342-46.
  12. Gibbs PA. Novel uses for lactic acid fermentation in food preservation. J Appl Bacteriol Symp Suppl 1987; 16:515-85.
  13. Beck C, Necheles H. Beneficial effects of administration of Lactobacillus acidophilus in diarrhea and other intestinal disorders. Am J Gastroenterol 1961;35:522-33.
  14. Tomic-Karovic K, Fanjek JJ. Acidophilus milk in therapy of infantile diarrhoea caused by pathogenic Escherichia coli. Ann Pediatr 1962;199:625-34.
  15. Gilliland SE, Speck ML. Antagonistic action of Lactobacillus acidophilus toward intestinal and foodborne pathogens in associative culture. J Food Protect 1977;40:820-23.
  16. Hosono A, Yastuki K, Thoita F. Isolation and characterization of an inhibitory substance against Escherichia coli, produced by Lactobacillus acidophilus. Milchwissenschaft 1977;32:727-30.
  17. Hamdan IY, Mikolajcik EM. Acidolin: an antibiotic produced by Lactobacillus acidophilus. J Antibiot 1974;27:631-36.
  18. Alm L. The effect of Lactobacillus acidophilus administration upon survival of Salmonella in randomly selected human carriers. Prog Food Nutr Sci 1983;7: 13-17.
  19. Gilliland SE. Beneficial interrelationships between certain microorganisms and humans. Candidate microorganisms for use as dietary adjuncts. J Food Protect 1979;42: 164-67.
  20. Cooke RD, Twiddy DR, Reilly PJA. Lactic acid fermentation as a low-cost means of food preservation in tropical countries. FEMS Microbiol Rev 1987;46:36979.
  21. Adams MR, Hall CJ. Growth inhibition of foodborne pathogens by lactic acid, acetic acid and their mixtures. Int J Food Sci Technol 1988;23:287-92.
  22. Reddy NR, Roth SM, Eigel WN, Pierson MD. Food and food ingredients for prevention of diarrhoea disease in children in developing countries. J Food Protect 1988;51:66-75.
  23. Axelsson LT, Chung TC, Dobrogosz WJ, Lindgren S. Production of a broad spectrum antimicrobial substance by Lactobacillus reuteri. Microbiol Ecol Health Dis 1989;2:131-36.
  24. Nout MJR, Hautvast GJ, van der Haar F. Marks WEW, Rombouts FM. Formulation and microbiological safety of cereal-based weaning foods. In: Alnwick D, Moses S, Schmidt OG, eds. Improving young child feeding in eastern and southern Africa: household-level food technology. Proceedings of a workshop held in Nairobi, Kenya, 11-16 Oct 1987. Ottawa. Ont, Canada: IDRC, 1988:245-60.
  25. Nout MJR, Rombouts FM, Hautvast GJ. Accelerated natural lactic fermentation of infant food formulations. Food Nutr Bull 1989;11(1):65-73.
  26. Mensah PPA, Tomkins AM, Drasar BS, Harrison TJ. Effect of fermentation of Ghanaian maize dough on the survival and proliferation of 4 strains of Shigella flexneri. Trans Roy Soc Trop Med Hyg 1988;82:635-36.
  27. Mensah PPA, Tomkins AM, Drasar BS, Harrison TJ. Fermentation of cereals for reduction of bacterial contamination of weaning foods in Ghana. Lancet 1990; 336: 14043.
  28. Svanberg U. Sj÷gren E, Lorri W. Svennerholm A-M, Kaijser B. Inhibited growth of common enteropathogenic bacteria in lactic fermented cereal gruels. World J Food Microbiol Biotech 1992;8:601-06.
  29. Chauliac M, Raimbault A-MM. Children in the tropics. Nutritional status: the interpretation of indicators. Review no. 181/182. Vincennes, France: International Children's Centre, 1989.
  30. Wilkinson L. SYSTAT: the system for statistics. Evanston, Ill, USA: SYSTAT, Inc. 1990.
  31. Winkelstein A. Lactobacillus acidophilus tablets in the therapy of various intestinal disorders: preliminary report. Am Pract Dig Treat 1955;1:1637-39.
  32. Gordon D, Macrae J. Wheater DM. A Lactobacillus preparation of use with antibiotics. Lancet 1957;1:889901.
  33. Niv M, Levy W. Greenstein NM. Yogurt in the treatment of infantile diarrhoea. Clin Pediatr 1963;2:407-11.
  34. Schmidt BJ, Ferreira JR, Ferreira M, Cury CP. Use of acidified whole milk in the prevention of infantile diarrhoea. Med Hyg 1976;34:276-77.
  35. Pearce JL, Hamilton JR. Controlled trial of orally administered Lactobacilli in infant diarrhoea. J Pediatr 1974;84:261-62.

Contents - Previous - Next