Contents - Previous - Next

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

Hunger, technology and society

Relating improvements in water supply and sanitation to nutritional status: The issue of using anthropometry as an evaluation measure
A field test for detecting iodine-enriched salt

Relating improvements in water supply and sanitation to nutritional status: The issue of using anthropometry as an evaluation measure

Raymond B. Isely

Water and Sanitation for Health Project, Public Health Research Physician, Research Triangle Institute, University of North Carolina Chapel Hill. North Carolina. USA


The interest in elucidating the relationship between water supply and sanitation improvements and nutritional status is more than an esoteric one. The International Drinking Water Supply and Sanitation Decade of the United Nations and the water supply and sanitation plan of the Alma Ata platform on "Health for All" are based on the assumption that investments in water supply and sanitation will improve the health of the populations served. Detractors from this view can point to the scarcity of evidence to support the assumption. The World Bank, in preparing for the decade, chose to avoid the question altogether by declaring that to do adequate research on the subject would cost more than anyone would care to invest (1). Yet programme managers of water and sanitation in this decade need to evaluate their programmes. Knowing the numbers of pumps installed, even numbers of functioning pumps, does not suffice. Investors and policy-makers demand evidence of health and other benefits as a justification for continued investment. Health benefits are among the first sought.

Among potential assessments of the health status of populations subjected to improvements in water supply and sanitation, measurements of nutritional status (primarily anthropometry) present a relatively easily managed, objective, and potentially sensitive reflector of improvements, on condition that the relationship to these improvements can be theoretically established. When contrasted with measures for the control of specific diseases, including prevalent conditions such as dysentery and gastroenteritis, anthropometry presents distinct advantages (2). Thus, the potential of using it in field evaluations of water supply and sanitation projects is strong,

Clarifying the relationship would also have a salutary effect on research on this subject. Those of us who have worked in the field have long realized the complexity of the relationships of nutritional status and disease incidence and prevalence with behaviour and environmental conditions. Influences flow in both directions along most of the links among variables. Each group of variables can be dissociated into hundreds of factors. There is need for more knowledge on how these mechanisms interact and interrelate.

Lastly and perhaps most important, a confirmed and better understood relationship is probably essential for planning at both policy and field levels within programmes in several sectors. Water supply and sanitation improvements could be a part of nutrition policy and strategy. Domestic use of irrigation water is a fact of life in most areas (3). Low-cost, simple adaptations of irrigation systems to address this reality could be an important adjunct to food production. An improved understanding of the links between water and nutritional health could lead to training in water supply and sanitation technology for field nutrition workers, and water and sanitation could be included in nutrition education programmes, as suggested by Latham (4) and others. Realizing those implications depends on how we address the relationship now. What is our present understanding? What evidence can be brought forth to support the various links in the chain of causation and correlation? What can we recommend now to evaluators, researchers, and programme managers?


A tentative model for understanding one pathway in the relationship between water supply and sanitation improvements and nutritional status is found in figure 1 (5). The model depicts a two-step quasi-causal or correlative relationship, with diarrhoeal morbidity intervening between the two. Intermediate between the improvements and diarrhoeal morbidity is a series of behavioural variables related to how facilities are accepted, used, and maintained. These are, in turn, later influenced by the degree and type of community participation. Impinging on the link between diarrhoeal morbidity and nutritional status is the occasional use of oral rehydration therapy, said by some to mitigate the nutritional effects of diarrhoeal disease (6). Moderating the response of nutritional status are food availability, represented by income, and feeding behaviour. Each of the elements in the diagram merits a brief discussion.

FIG. 1. Relationships among Improved Water Supply and Sanitation, Diarrhoeal Morbidity, and Nutritional Status (adapted from Isely, ref. 5)

Improved Water Supply and Sanitation

One must assume that systems, however simple, are correctly installed. Without such an assumption further discussion of these relationships would be futile. Another assumption is that systems produce water in sufficient quantity and quality and are reasonably convenient and reliable. The World Health Organization (7) has established standards for quality and convenience of water. Hughes (8), in reviewing 43 studies of varying quality, has concluded that projects that emphasized increased quantities of water were more frequently able to demonstrate improvements in diarrhoeal morbidity than those that emphasized only improved water quality. An accepted minimum per capita figure necessary for health benefits appears to be 20 to 30 litres a day (8).

Regarding sanitation, the World Health Organization (7) has not ventured further than declaring that facilities must be "adequate." Kalbermatten et al. (9) have published a series of standards for various latrine constructions. In practice, however, these models have proved to be more costly than originally estimated. The simple pit-latrine that the World Bank publication eschews may still be an "adequate" option.

Standards for personal and food hygiene are obviously difficult to set in rural or pert-urban areas of developing countries. Even in urban areas, application of food hygiene standards is spotty at best. We are thus left with the dilemma of defining what is meant by "improved" water supply and sanitation. Since there is no simple way out of the dilemma, one must assume a certain level of improvements, either meeting standards already set or those that appear adequate to help reap the health benefits we are seeking.

Acceptance, Appropriate Use, Maintenance, and Repairs

Acceptance of new water supply technologies appears to depend chiefly on convenience and reliability (10). Appropriate and careful use of facilities can be assured by user education, dependent in turn on community involvement in collective behavioural change (11). Maintenance and repair of water supplies are likewise dependent on community participation, but conditioned strongly by the availability of technical assistance and spare parts.

The acceptance, use, and maintenance of sanitation installations sufficient to assure health effects appear to be a response to efforts at user education. Community participation is probably even more essential. Shuval et al. (12) contend that this response threshold probably cannot be overcome in the absence of a certain level of socio-economic development. However, their model ignores the possibility that a sanitation education programme could be sufficiently intensive.

Diarrhoeal Morbidity

Diarrhoeal morbidity has proved extremely difficult to define. Frequency of diarrhoeal episodes is an elusive concept, especially in circumstances where diarrhoea tends to be a chronic, continuing phenomenon. Rowland et al. (13, 14), for example, estimated that the average rural Gambian infant spends 20.4 per cent of the rainy season with diarrhoea. In such circumstances, the boundary between one episode and the next tends to be blurred. This problem confounded the evaluation of the Guatemala rural water supply project as reported first by Schneider and co-workers (15) and later by Dworkin and Dworkin (16). Diarrhoeal episodes in the latter reworking of the data were treated as discrete variables, whereas they are probably on a continuum.

Frequency of stools of a diarrhoeal type represents a second way to measure diarrhoeal morbidity, but problems of definition tend to render data gathering unreliable. Rahaman (17) in Bangladesh is using pictorial presentations of diarrhoeal stools in interviewing mothers to overcome some of these problems.

As indicated in the diagram, diarrhoea is not a single entity; rather, it is a syndrome composed of conditions with multiple aetiologies. Rahaman and Wahed (18) have demonstrated differing degrees of protein and calorie loss for the various aetiologic agents associated with diarrhoeal disease. It appears that Shigella species, rotavirus, and perhaps Escherichia coli are the chief offenders. The key factor appears to be the ability of the organism to penetrate the epithelial layer of the gut (19). Bacteriologic or virologic diagnosis as a basis for assessing diarrhoeal morbidity is not, however, very promising under most field conditions. (Carrying this out in the evaluation of the Imo State-UNICEF Water Supply and Sanitation Project required special arrangements for air shipment of specimens to a university laboratory [20].) Neither can one ignore the possibility that a group of infants and children would have diarrhoea of many aetiologies in a given time frame. A more reliable and more feasible indicator is needed.

FIG. 2. Model of the Influence of Improved Water Supply (from L. Chen, "Evaluating the Health Benefits of Improved Water Supply through Assessment of Nutritional Status in Developing Countries," unpublished paper, Harvard School of Public Health, Boston, 1980)

Chen (unpublished data, 1980) suggests nutritional status as an alternative indicator of a good effect of improved water supply. His model (fig. 2) depicts the influence of improved water supply and sanitation as mediated through not only diarrhoeal morbidity but also through savings in maternal time and energy and improved agricultural production, which are passed on to the infant. The energy savings may be most critical to the breast-feeding women (21, 22). In other circumstances, diarrhoeal morbidity is probably the most important intermediate. In fact, diarrhoeal disease, by causing intestinal losses, increased catabolism, and anorexia, may be the chief reason for growth failure.

Nutritional Status

Measuring growth has the advantage of relative ease of administration by personnel with limited training under various conditions in the field. Interpretation of growth data is aided by the existence of standards for height for age, weight for age, weight for height, and arm circumference. Each measurement has advantages and disadvantages that are well known (23).

Mediating Variables: Oral Rehydration, Feeding Habits, and Income

Among the most salient objections to the use of nutritional status as an indicator of the effectiveness of water supply and sanitation programmes are the possible influence and/ or interference of oral rehydration, feeding habits, income, and other mediating variables. Oral rehydration is of unquestioned value in preventing death from severe diarrhoea, but the extent to which it prevents the adverse nutritional effects of diarrhoeal disease is uncertain. Hirschhorn and Denny (6) report that Apache children experiencing a rapid restoration of body fluids through the use of an oral glucose-electrolyte solution tolerate oral feedings sometimes within hours of restoring equilibrium. Those children who also received oral feedings were discharged with significantly higher median weights than those who did not.

In a similar study of Philippine children, Hirschhorn and an international study group (24) report a salutary effect on weight gain in children over one year of age receiving an oral glucose-electrolyte solution and continuous feeding.

Weight gains were significantly different from those of controls for one episode, one month after an episode, and seven months after an episode. Rowland and Cole (25), however, in a controlled study could not demonstrate any significant effect.

Since any changes in infant and child feeding habits or household income during the period in which water supply and sanitation installations were being evaluated might significantly affect the nutritional status of the infant and child population and thereby lead to a confused interpretation of results, these factors need to be controlled for when programmes are being evaluated. Feeding habits might be expected to change in the presence of an intensive nutrition education programme. Income, a proxy for food availability, might rise if programmes of rural credit, irrigation, fertilizers, or new high-yield varieties of food were successful.

The foregoing section has presented a model depicting the relationships among improved water supply and sanitation, diarrhoeal morbidity, and nutritional status. Each of the variables has been characterized, and moderating variables have been described. Now the evidence for the linkages among them will be examined as a prelude to some concluding recommendations


Evidence Supporting the Linkage of Improved Water Supply and Sanitation to Diarrhoeal Morbidity

The linkage between improved water supply and sanitation and diarrhoeal morbidity is somewhat better supported than one might suppose. The weaknesses in many studies are probably related to the difficulties alluded to earlier of measuring diarrhoeal morbidity in the field and to problems associated with the use of installed facilities (26, 27). Hughes (8) and McJunkin (28) have recently reviewed the evidence supporting a positive influence of improved water supply on diarrhoeal morbidity. As indicated earlier, Hughes, in reviewing some 43 studies, concluded that there is more evidence than one might expect for a positive influence. Of interest is his observation that an increased quantity of water is more beneficial than increased quality of water to achieve a diminution in diarrhoeal morbidity. Unfortunately, the studies reviewed were not all of the same quality. Some were poorly designed and/or executed. Hughes admitted studies to his review if indicators of diarrhoeal disease in populations exposed to different levels of water supply and/or sanitation were quantified. All studies had to contain data for at least one study and one control group. Excluded were epidemic investigations, case-control studies, studies where the different water supply and/or excrete disposal groups were not discrete, or studies where results were presented as linear regressions or correlations without adequate data to permit calculations of indicator rates in study and control populations.

McJunkin (28) has recently reviewed several hundred studies relating water supply and sanitation improvements to health indicators, the majority of the latter those of enteric disease. He concludes that there is a health impact, especially where water is readily accessible in adequate quantities. Other things being equal, health impacts improve with increasing levels of service and with complementary activities in excrete disposal and health education. Soap appears essential. Interestingly in this review, the studies that were unclear or negative tended to occur in situations of the greatest poverty and under development

The first major linkage in the model is thus supported by the weight of evidence, both historical and contemporary.

Evidence Supporting the Linkage of Diarrhoeal Morbidity to Nutritional Status

The relationship between diarrhoeal morbidity and nutritional status is probably the best supported of the three (see 29 and 30). There is little need to review individual studies. Table 1 summarizes several of the more important of the past several years (31 - 39). As the table indicates, the relationship of diarrhoeal morbidity to physical growth is stronger than for any other disease, and probably operates with shortages of food to keep growth rates below standard for the majority of poor children despite medical and nutrition interventions. Other factors, too, seem to condition the response of growth rates to diarrhoeal disease: sex, the parity of the mother, caste or socio-economic level, and the timing of birth and weaning in relation to season.

Evidence Supporting the Linkage of Improved Water Supply and Sanitation to Nutritional Status

Whether through diminished morbidity, savings in maternal time and energy, or improved food production, a number of studies are beginning to suggest nutritional benefits from improvements in water supply and/or sanitation. Although there is great variation in the quality of these investigations, results are encouraging enough to merit attention. Those addressing the links between improvements in water supply and sanitation and nutritional status are summarized in table 2 (40-42).

The results of a baseline study by Anderson (40) are mostly speculative. Carefully collected baseline nutritional indices revealed a predominance of stunting in rural Bolivian (Quecha) preschool children. Stunting was more pronounced at higher than at lower altitudes. Wasting was more frequent at lower altitudes. Dietary analysis revealed lower caloric and protein content at higher altitudes; gastroenteritis was more frequent at lower elevations. All deficits were most pronounced in the second year of life.

TABLE 1. Studies or Reviews of Studies Concerning the Effect of Diarrhoeal Morbidity on Nutritional Status

Year Author(s) Location of Study No. of Subjects Independent Variable(s) Nutritional Variable(s) Observations
1961 McGregor et al. (31) Gambia 195 cohort Season Height Coincidence of growth faltering, infection, and mortality in rainy season associated with increased maternal agricultural activity
Infections Weight
Food availability
Preventive and curative care
1968 Guzman et al. (32) Guatemala 816 cohort Feeding Height Feeding village children showed slightly better growth than children in villages with environmental improvements, but still far below standards
Environmental improvements Weight
Preventive and curative care Bone development
Head circumference
1972 Mata et al. (33) Guatemala 45 cohort Infection with enteric viruses Weight Nutritional deficits from viral infection in first 3 months persist to 12 months. Growth deceleration associated with:
Recuperation slow
Head and thoracic circumference
1975 Martorell et al. (34) Guatemala 716 cohort Morbidity:
—respiratory illness
Growth in height and weight Negative correlation with diarrhoea only
No significant interactions
1976 McGregor (35) Gambia 258 cohort Season Weight Variation with seasons, infection rates, and weaning
Timing of birth Height
Food availability Weight/height
Malaria Haemoglobin Varies with season
1977 Rowland et al. (36) Gambia ? Gastroenteritis Height/age Significant relationship to gastro enteritis; for malaria, to weight only
Malaria Weight/age
1977 Cole and Parkin (37) Gambia and Uganda 197 Infectious morbidity
Weight Growth significantly related to gastroenteritis
(P < .001) and to malaria (P < .01)
Effect of gastroenteritis vs. malaria = 13:1
1978 Taylor et al. (38) and, Kielmann et al. (39) India 2,900 cohort Sex, caste Height Nutrition care, nutrition and medical care combined, caste, and sex equally influenced anthropometric variables
Nutrition care Weight
Preventive care
Parity of mother Up to 50% remained below Harvard standards
No. of male children
1979 Rowland (unpublished) Gambia and Uganda 152 Diet Weight-height/age Comparison of Ugandan and Gambian children: Ugandan growth > Gambian unrelated to diet; related to gastroenteritis

What possible role does improved water supply and sanitation have in improving the nutritional status of these infants? One would assume that water supply and sanitation improvements influence to a far greater extent the wasting found predominantly at lower altitudes than the stunting more prevalent at higher altitudes. Although diarrhoeal disease contributes to wasting mainly through caloric deprivation, one cannot ignore possible influences of diarrhoea on chronic malnutrition through morphologic and physiologic changes in intestinal mucosa and resulting malabsorption. Thus, improvements in water supply and sanitation practices might be expected to influence nutritional status at all altitudes, although to a greater extent at lower altitudes where diarrhoea and wasting are more prevalent. The role of genetic predisposition to short stature at higher elevations may be important, although as Anderson points out, Quecha children in Peru have achieved a median height of 25 per cent of standards when subjected to supplementary feeding.

TABLE 2. Studies of the Relationship of Improved Water Supply and/or Sanitation to Nutritional Status

Year Author(s) Location of Study No. of Subjects Independent Variables Nutritional Variable(s) Observations
1978 Tomkins (42) Nigeria 183 Water supply
—protected copious
—unprotected scanty
Height Weight/age and weight/height greater in villages with copious protected water (P < .01 )
Midarm circumference
Food hygiene
1981 Anderson (40) Bolivia 221 Infection Weight Baseline data only
—majority with adequate weight for height
—majority with > 1 SD weight/age
Potable water systems Height
Breast-feeding and weaning practices  
Diet 24-hour dietary history
Altitude   Diets 68% caloric deficient; 30% protein deficient; 18% current diarrhoea rate; 35% past diarrhoea rate (38% in those < 3 yrs.; 47% < 2 yrs.) girls > boys
    Diets were more deficient at high altitudes; diarrhoea more prevalent at lower altitudes
1981 Henry (41) St. Lucia 75 cohort Water supply Height Height and weight increments improved in areas with household taps and latrines
Latrines at different levels Weight
Controls for literacy, house construction, cooking facilities, garbage disposal, infant food preparation

The studies of Henry (41) and Tomkins et al. (42) are more striking. In both there were decidedly significant differences in weight-for-age and weight-for-height of children benefiting from sanitary improvements.

The Henry study is particularly noteworthy because of its longitudinal design. Water supply (household taps) was the critical variable. Although having a latrine as well did not produce a better result in anthropometric findings, results for prevalence of diarrhoea and intestinal helminths were significantly better. Interpretation of the study is hampered by the small sample size. Growth in children initially retarded also tended to catch up after the third year of life. It could be asked how many in the group with early retarded growth survived until the third year.

Tomkins and colleagues (42) compared children in villages having "scanty, unprotected" with those having "copious, protected" water supplies. One wonders if the beneficial effect they noted on weight increments was due to the protected quality of the water or to its abundance. It is noteworthy that both copious and scanty water supplies were from open wells. The former, however, had parapets and drainage aprons and even greater flows and greater accessibility than the more traditional wells. The greater quantities of available water, with implications for personal and food hygiene, offer a more plausible explanation for the results.

This limited search has not produced a single study demonstrating a direct connection of improved water supply with savings in maternal time and energy or increased food production to improved nutritional status of children. Studies of the consequences of mothers' agricultural activities are nonetheless instructive. Thompson and Rahaman (43) and McGregor et al. (31 ) some time ago described the results for children unfortunate enough to arrive at weaning at the beginning of the crop-growing season in the Gambia. If their mothers' rice fields were at some distance from the village, there was a tendency to wean abruptly and to leave the child with a younger sibling or an elderly person in the family. These children had a demonstrably higher infection rate, a greater degree of weight faltering, and a higher mortality rate than those who passed through the growing season at the breast.

Vis et al. (22) have shown a similar pattern for women in north-east Zaire. Those travelling the greatest distances to fields not only had children with the highest infection rates, earlier weight faltering, and higher mortality, but the mothers themselves had the least weight gain in pregnancy, the smallest newborns, and the smallest volumes of breast milk. Other variables of importance were season, parity, and maternal infections.

By extrapolation, these same consequences for the nutritional status of infants and children could result from long distances to water coinciding with critical times in the cycle of pregnancy, lactation, and child growth and development. It has been reported (21 ) that in hilly terrain a woman may consume 27 per cent of her daily energy intake. In the lactating woman such an expenditure would leave a very small caloric reserve.


Evidence has been reviewed for the relationship of improved water supply and sanitation to nutritional status in young children. A two-step hypothetical model of causation has been examined. Each step, as well as the overall equation, has been weighed in the light of available evidence in field studies. There seems to be sufficient evidence to support each linkage to a variable extent. The water supply/sanitation/diarrhoeal morbidity linkage has moderate support. That between diarrhoeal morbidity and nutritional status is strongly supported, whereas the overall relationship between water supply/sanitation and nutritional status remains tentative, but promising. The consequences of this analysis for evaluation, research, and policy need to be carefully examined.

Evaluation of Water Supply and Sanitation Programmes

The question before evaluators is whether to recommend nutritional status as an evaluation indicator now, or to await further research. At least two important research studies at Teknaf, Bangladesh (17), and at Jhansi, India (44), are in progress. In both, anthropometry is a major dependent variable in what amounts to a four-cell research design where villages receive varying combinations of water supply, sanitation, and health education. Investigators in several water supply and sanitation projects or involved with components of other projects in primary health care or rural development that have been underway for one or more years are considering the use of anthropometry as an evaluation indicator. Should this step be recommended?

The relative ease of taking measurements, their reliability in the hands of minimally trained personnel, their relatively low cost, and the evidence supporting the relationship, especially the studies of Henry (41) and Tomkins (42), all combine to support the recommendation that anthropometry be used in this way. As a consequence of such reasoning, researchers in the Philippines, Togo, Somalia, and Lesotho are planning such a move. The cumulative results of these evaluations may be as useful as the results of the most perfectly designed research.


Research is needed to clarify the role of increased quantity versus improved quality of water with respect to nutritional status. If increased quantity is the key factor, then how does it operate? There is evidence, for example, that contaminated foods, particularly weaning foods, may be the principal mechanism for transmitting diarrhoea and dysentery (14, 45, 46). How does an increased quantity of household water help to arrest this transmission? What is the role of excrete disposal? The influences of such immutable factors as caste or socio-economic level, altitude, and season need further exploration as to their relative importance in determining the end result of improved growth of children. Also, further evidence is needed for the role of oral rehydration and early feeding of children with diarrhoea in mitigating chronic malnutrition of young children.

Lastly, there is the question of how the multiple ways in which maternal savings in time and energy brought about by shorter distances and fewer difficulties in obtaining water can contribute to better child nutrition.


Should water supply and sanitation improvements be viewed as part of a national nutrition policy? Chen would respond affirmatively. He views water primarily as a nutrient and improvements in water supply accessibility, reliability, and quantity as having consequences primarily in the improved growth of children. Whether one agrees or not, the value of associating water supply and sanitation and nutritional programmes in the field seems obvious. Given that the role of diarrhoea is at least equal to that of food availability among the determinants of nutritional status, effective means of diarrhoeal disease prevention need to be considered.

Policy and programme changes should probably await the results of current research and evaluation efforts, however, in order to gather adequate support. The evidence in favour of such moves is growing but is not yet satisfactory.


1. Public Utilities Department, World Bank, "Measurement of the Health Benefits of Investments in Water Supply," Public Utilities Report no. PUN 20 (World Bank, Washington, D.C., 1 976).

2. R. Struba and R, B. Isely, "The Choice of Health Status Indicators to Evaluate Water and Sanitation Projects in North Cameroon," Water and Sanitation for Health Project Technical Report no. 5 (Arlington, Va., USA, 1982),

3. S. Annis and S. Cox, "The Integration of Small-Scale Irrigation and Village Potable Water Systems in Guatemala," Water Supply and Management, 615): 455 (1982).

4. M. Latham, "Strategies to Control Infections in Malnourished Populations-Holistic Approach or Narrowly Targeted Interventions?" Amer. d. Clin. Nutr., 31: 2292 (1978).

5. R. B. Isely, ed, "Relationships among Improved Water Supply and Sanitation, Diarrheal Morbidity, and Nutritional Status: Report of a Working Meeting," Water and Sanitation for Health Project Working Paper no. 14 (Office of Health, Bureau for Science and Technology, USAID, Washington, D.C., 1982).

6. N. Hirschhorn and K. M. Denny, "Oral Glucose Electrolyte Therapy for Diarrhea: A Means to Maintain or Improve Nutrition," Amer. J. Clin. Nutr. 28: 189 (1975).

7. World Health Organization, "Community Water Supply and Wastewater Disposal, Mid-decade," Wrld. Hlth. Statist. Rep., vol. 29, no. 10 (WHO, Geneva, 19761.

8. J. Hughes, "Potential Impacts of Improved Water Supply and Excreta Disposal on Diarrheal Disease Morbidity: An Assessment Based on a Review of Published Studies," (WHO, Geneva, 1980).

9. J. M. Kalbermatten, D. S. Julius, and C. G. Gunnerson, A Sanitation Field Manual, Appropriate Technology for Water Supply and Sanitation Series, vol. 11 (World Bank, Washington, D.C. 1980).

10. G. F. White, D. Bradley, and A. U. White, Drawers of Water (University of Chicago Press, Chicago, III., USA, 19721.

11. R. B. Isely and K. Parker, "Application of Health Education to Water Supply and Sanitation Projects in Africa," Water and Sanitation for Health Project Technical Report no. 15. (Office of Health, Bureau for Science and Technology, USAID, Washington, D.C., 1982).

12. H. I. Shuval, R. L. Tilden, B. H. Perry, and R. E. Grosse, "Effect of investments in Water Supply and Sanitation on Health Status: A Threshold-Saturation Theory.'' Bull. Wld. Hlth. Oral, 54 (2): 243 (1981).

13. M. G. M. Rowland, T. J. Cole, and R. G. Whitehead, "A Quantitative Study into the Role of Infection in Determining Nutritional Status in Gambian Village Children," Brit.. J. Nutr., 37: 441 (1977).

14. M. G. M. Rowland and P. K. McCollum, "Malnutrition and Gastroenteritis in the Gambia," Trans. Roy. Soc. Trop. Med. Hyg., 1 (3): 199 (1977).

15. R. E. Schneider, M. Shiffman, and J. Faigenblum, "The Potential Effect of Water on Gastrointestinal Infections Prevalent in Developing Countries." Amer. J. Clin. Nutr., 31: 2089 (1978).

16. D. Dworkin and J. Dworkin, "Water Supply and Diarrhea: Guatemala Revisited," Evaluation Special Study no. 2 (USAID, Washington, D.C., 1980).

17. M. M. Rahaman, "Research Protocol for Water and Sanitation Intervention Project, Teknaf, Bangladesh" (International Centre for Diarrhoeal Disease Research-Bangladesh, Dhaka, Bangladesh, 1980).

18. M. M. Rahaman and M. A. Wahed, "Direct Nutrient Loss in Diarrhoea," Glimpse (ICDDRB), 3 (10): 4 (1981).

19. S. B. Formal, H. L. Dupont, R. Hornick, M. J. Snyder, J. Libonati, and E. H. LeBrec, "Experimental Models of the Investigation of the Virulence of Dystentery Bacilli and Escherichia coli, " Ann N. Y. Acad. Sci., 176: 190 (1971) .

20. D. Blum, "Evaluation of the UNICEF-Assisted Imo State Water Supply and Sanitation Project: Description of Methodology" (The Ross Institute, London, 1983).

21. R.B. Isely, "The Relationship of Accessible Safe Water and Adequate Sanitation to Maternal and Child Health-Looking Forward to the Drinking Water and Sanitation Decade," Water Supply and Management 5 (6): 417 (19811.

22. J. L. Vis, P. Hennart, and M. Ruchababisha, "L'Allaitement en Zone Rurale Pauvre," Carnets de l'enfance, 55/56: 172 (1981).

23. J. P. Habicht, L. D. Meyers, and C. Brownie, "Indicators for Identifying and Counting the Unproperly Nourished," Amer. J. Clin. Nutr., 35: 1241 (1982).

24. N. Hirschhorn and an International Study Group, "Beneficial Effects of Oral Electrolyte Solutions in the Treatment of Children's Diarrhea," J. Trop. Pediat., 17:62 (1981).

25. M. G. M. Rowland and T. J. Cole, "The Effect of Early Glucose-Electrolyte Therapy on Diarrhoea and Growth in Rural Gambian Village Children," J. Trop. Pediat., 26: 54 (1980).

26. G.T. Curlin, K. M. A. Aziz, and M. D. Khan, ''The Influence of Drinking Tubewell Water on Diarrhoea Rates in Matlab Thana," Research Working Paper no. 1 (International Centre for Diarrhoeal Disease Research-Bangladesh, Dhaka, Bangladesh, 1979).

27. J. Briscoe, "The Role of Water Supply in Improving Health in Poor Countries (with Special Reference to Bangladesh)," Amer. J. Clin. Nutr., 31: 2100 (1978).

28. F. E. McJunkin, Water and Human Health (USAID, Washington, D.C., 1982).

29. N. S. Scrimshaw, C. E. Taylor, and J. E. Gordon, Interactions of Nutrition and Infection, World Health Organization, Monograph Series, no. 57 (WHO, Geneva, 1968).

30. M. Latham, "Diet and Infection in Relation to Malnutrition in the United States," N.Y. State J. Med., 15 Feb. 1970, p. 558.

31. I. A. McGregor, W. Z. Billewicz, and A. M. Thomson, "Growth and Mortality in Children in an African Village," Brit. Med. J., 12/23: 1661 (1961).

32. M. Guzman, N. S. Scrimshaw, H. A. Bruch, and J. E. Gordon, "Nutrition and Infection Field Study in Guatemalan Villages, 19591964. VII: Physical Growth and Development of Prschool Children," Arch. Environ. Hlth., 17: 107 (1968).

33. L. J. Mata, J. J. Urrutia, C. Albertazzi, O. Pellecer, and E. Arellano, "Influence of Recurrent Infections on Nutrition and Growth of Children in Guatemala," Amer. J. Clin. Nutr., 25: 1267 (1972).

34. R. Martorell, J. P. Habicht, C. Yarbrough, A. Lechtig, R. E. Klein, and K. A. Western, "Acute Morbidity and Physical Growth in Rural Guatemalan Children," Amer. J. Dis. Child., 129:1296 (1975).

35. I. McGregor, "Health and Communicable Disease in a Rural African Environment,', Oikos, 27: 180 (1976).

36. M. G. M. Rowland, T. S. Cole, and R. G. Whitehead, "A Quantitative Study into the Role of Infection in Determining Nutritional Status in Gambian Village Children," Brit. J. Nutr., 37:441 (1977).

37. T. J. Cole and J. M. Parkin, "Infection and Its Effect on the Growth of Young Children: A Comparison of the Gambia and Uganda," Trans. Roy. Soc. Trop. Med. Hyg., 71 (3): 196 (1977).

38. C. E. Taylor, A. Kielmann, C. DeSweemer, I. S. Uberoi, H. S. Takulla, C. G. Neumann, W. Blot, H. Shankar, S. Vohra, C. Subbulakshmi, R. S. S. Sarmar, R. L. Parker, C. McCord, N. Masih, D. Laliberte, N. S. Kielmann, D. N. Kakar, and A. Forman, "The Narangwal Experiment on Interactions of Nutrition and Infections. I: Project Design and Effects upon Growth," Indian J. Med. Pes., 68 (suppl.): 1 (1978).

39. A. Kielmann, C. E. Taylor, and R. L. Parker, "The Narangwal Nutrition Study: A Summary Review," Amer. J. C/in. Nutr. 31: 2040 (1978).

40. M. A. Anderson, "Health and Nutrition Impact of Potable Water in Rural Bolivia," J. Trop. Pediat., 27:39 (1981).

41. F. Henry, "Environmental Sanitation and Nutritional Status of Infants in Rural St. Lucia, West Indies," Trans. Roy. Soc. Trop. Med. Hyg., 75 (4): 507 (1981).

42. A. M. Tomkins, B. S. Drasar, A. K. Bradley, and W. A. Williamson, "Water Supply and Nutritional Status in Rural Northern Nigeria," Trans. Boy. Soc. Trop. Med. Hyg., 72 (3): 239 (1978).

43. B. Thompson and A. K. Rahaman, "Infant Feeding and Child Care in a West African Village," J. Trop. Pediat., 13: 124 (1967).

44. R. N. Srivastava, "Protocol for the Study on the Health Benefits of Water Supply in a Rural Area in Uttar Pradesh." (Jhansi Medical Colleqe, Jhansi, India, 1981).

45. E. Capparelli and L. J. Mata, "Microflora of Maize Prepared as Tortillas," Appl. Microbiol., 29: 802 (1975).

46. W. Smith, "Media and Health Practices: Results of Honduras Investigation," Academy for Educational Development Report no. 14 (Washington, D.C., 1980).


A field test for detecting iodine-enriched salt

J.-P. Dustin
Chief, Food Aid Programmes, Division of Coordination, World Health Organization, Geneva, Switzerland

J.-P. Ecoffey
Pharmacist, Geneva, Switzerland


The following article presents a practical means of monitoring compliance with regulations for the iodization of salt for the prevention of endemic goitre. It is reprinted here because the method has received insufficient attention and its availability is not widely known. Endemic goitre remains a serious public health problem in many countries despite the feasibility of its prevention through salt iodization. The availability of this inexpensive, rapid, and reliable means of testing for added iodine should help the adoption of salt-iodization legislation where needed.

The iodization of salt at some appropriate level is now the most widely used prophylactic public health measure against endemic goitre. While industrialized countries stipulate potassium (or sodium) iodide, developing countries usually prefer potassium iodate, which is more expensive but keeps better in storage, particularly in humid or warm conditions or when exposed to light.

If the locally recommended iodization agent is not known, two samples should be examined, one by the iodide test and the other by the iodate test. For routine controls, the simplest approach is to compare the unknown salt with a standard sample of salt iodized according to local regulations, using the relevant test only.


The test described below will detect the presence of the iodide ion over the range of officially recommended levels of iodization (5-100 mg of potassium iodide per kilogram of salt).


The following three solutions should be prepared:

Solution A. 50 ml of a 0.5% starch solution, made by boiling 50 ml of water with 0.25 9 of rice starch for one minute and letting it cool. The resulting liquid is whitish, contrary to the normal colourless laboratory iodometric starch solutions; however, this is irrelevant to the proposed test.

Solution B. 25 ml of 1% sodium nitrite (0.25 9 in 25 ml of distilled water). 0.25-9 capsules of dry crystalline nitrite can be provided if this is found useful.

Solution C. 25 ml of a 20% solution of 95% sulphuric acid (specific gravity 1.83),

All three solutions should be stored in glass-stoppered dropper bottles.

The iodide reagent is obtained by mixing 50 ml of solution A, 10 drops (0.5 ml) of solution B, and 10 drops (0.5 ml) of solution C. It is stable for two to three days under temperate laboratory conditions.


The test is carried out as follows: On a saucer, place a small amount of the salt to be tested and, separately on the same saucer, a similar amount of salt iodized at the locally legal level. Moisten both portions of salt with two drops of the reagent. The wet iodized salt should turn blue immediately, and the colour will remain visible for several minutes before turning grey and eventually white (after about 30 minutes). If the salt being tested also turns the same blue, it is properly iodized.


This test cannot be used to measure the relative degree of iodization in different samples because it produces a

uniformly light blue colour over much of the official range of concentrations. Also, it cannot be used to detect the iodate ion because the reagent does not react with iodates in a visible way. It works best on the usual finely crystalline salts; it is less sensitive with very finely ground salts which are not wetted by the reagent as readily as the crystalline forms.


This test will detect the presence of the iodate ion over the range of officially recommended levels of iodization (6-130 mg of potassium iodate per kilogram of salt).


The following three solutions should be prepared:

Solution I. Solution I is identical to solution A above.

Solution II. 50 ml of 12% potassium iodide (6 9 of Kl in 50 ml of distilled water).

Solution III. 25 ml of 10% hydrochloric acid (specific gravity 1.05). If necessary, it may be prepared by mixing 10 ml of concentrated HCI (specific gravity 1.13) with 15 ml of distilled water.

All three solutions should be stored in glass-stoppered dropper bottles.

The iodate reagent is obtained by mixing 25 ml of solution II, 12 drops of solution III, and 25 ml of solution I. It is stable for two to three days under temperate laboratory conditions.


The test is carried out as follows: On a saucer, place a small amount of the salt to be tested, and separately on the same saucer, a similar amount of salt iodized at the locally legal level. Moisten both portions of salt with two drops of the reagent. The wet iodized salt should turn grayish-blue immediately, and the colour will remain visible for several minutes before turning brown. If the salt being tested turns the same grayish-blue, it is properly iodized.


This test can be used to estimate roughly the relative degree of iodization in different samples because it produces some range of grayish-blue colour over much of the official range of concentrations. It cannot be used to detect the iodide ion because solution II is itself iodide: the test needs iodide to free iodine from iodate in a visible manner.

Travelling Kit

A travelling kit can conveniently be assembled. For the iodide alternative it would consist of one 60-ml dropper bottle, two 30-ml dropper bottles, and two 50-ml powder flasks to contain, respectively, solution A (and the completed reagent), solutions B and C, powdered starch (or unguent), and iodized salt. For the iodate alternative the same kit would suffice if solutions II and III were substituted for solutions B and C respectively. A kit allowing the detection of both iodide and iodate would require four 30 ml dropper bottles to contain solutions B, C, II, and III. It is recommended that stoppers be secured with string tied in a pharmacist's knot. Measuring spoons, one 50-ml graduated cylinder, one white-glazed ceramic tile, and two glass-rod stirrers may be convenient additions.


1. If no standard balance is available, weighings may be carried out on a good letter balance. Volumes of liquids may be measured with graduate cylinders and "drops" are those delivered by standard medicine droppers (about 0.05 ml each). Such equipment, if carefully used, is sufficiently accurate for these tests.

2. If it is impractical to boil the starch, a suitable alternative is "unguentum glycerini," which is a standard preparation in several pharmacopoeias. it is prepared from, in parts by weight: wheat starch, 10; water, 15; and glycerol, 90. The wheat starch and water are mixed to a homogeneous suspension, the glycerol is then added, and the mixture is warmed over a water bath (90 C) until it becomes uniformly translucent in thin layers. Under tropical conditions, this unguent may become mouldy over a period of days or weeks; however, if the surface layer of mould is removed, the remaining mixture may still be used. If mould growth becomes a serious problem, 6 parts per thousand of thiomersal, incorporated after the glycerol, will act as an effective preservative. An alternative preservative consists of 0.1 9 of methyl-p-hydroxybenzoate per 100 9 of unguent; although this preservative has not been tested under tropical conditions, it is effective in temperate climates and does not interfere with the proposed spot tests. Half a teaspoonful of the unguent plus 45 ml (3 tablespoonfuls) of water, well mixed to a homogeneous whitish suspension, provides a good alternative to solution A.

3. In tropical climates the solution will tend to become mouldy. It will keep longer if 5 g of thiomersal powder are added to each 25 g of starch, which gives a final concentration of 0.1% thiomersal in the solution. There is little doubt that other preservatives may be used instead, their choice being a matter of local availability, effectiveness, and lack of interference with the spot tests themselves.

4. Under tropical conditions the least stable component is solution A, and it is the solution that will need to be prepared anew if the positive test fails. Solution B should be the second component to be suspected, while solution C is adequate as long as it remains colourless.

5. Under tropical conditions the least stable component is solution I, and it is this solution that will need to be prepared anew if the positive test fails. Solution II should be the second component to be suspected. Solutions II and III are adequate as long as they remain colourless.

Contents - Previous - Next