HomeLibrary5-a-DayFAQsEspañolLeaf ConcentrateYou Can Help 

The Effect of a Leaf Concentrate Supplement on Haemoglobin Levels in Malnourished Bolivian Children: A Pilot Study

Ceri A. Lowe.


Waterlow (1962) was amongst the workers showing an interest in Leaf Concentrate (LC) and showed its value as a protein supplement. Produced locally from green leaves, LC is a protein curd The socalled 'Protein Gap, of the 1960s led to a search for alternative protein sources. which also contains approximately 1mg of nonhaem iron per gram. (See Appendix: 'Approximate Composition of LC,).

Although much of the work involving LC has considered LC as a nitrogen source (Shah et al, 1981); it has been suggested by Olatunbosun et al (1972) that it may reduce the incidence of folic acid and iron deficiency related anaemias. It is not known if the giving of vitamin C with LC enhances this effect, although the increased absorption of nonhaem iron in the presence of vitamin C is well documented (Chanarin, 1988). It is known that the iron present in LC is more readily absorbed than from the leaves from which it is extracted as the leaf fibre is removed during processing; this processing removes most of the vitamin C present in the original leaves.

Anaemia is said to affect 49.7% of rural preschool children in Bolivia and iron deficiency anaemia is the most prevalent (UNICEF 1989). Anaemic children are smaller and grow at a slower rate than their nonanaemic counterparts (Chwang et al, 1988). Mental development may be retarded and attention span reduced in such children (Pollitt et al, 1986). The symptoms of iron deficiency are more pronounced at high altitude and are apparent at higher haemoglobin levels than at sea level.

The aim of this study was to evaluate quantitatively the efficacy of using LC, with and without vitamin C, as a means of treating irondeficiency anaemia in malnourished Bolivian children. Such effects of this supplementation as changes in height, weight and the incidence of common illnesses were also monitored.


Recruitment of Subiects.

Forty five preschool children were randomly selected from community health records in the locality of Machacamarca, a village 30kms from Oruro on the Bolivian Altiplano, at approximately 3800 metres above sea level. All the children were either first or second degree malnourished i.e. with a weight for age between 60 and 90% of the World Health Organisation (1983) 50th centile. The 25 male and 20 female children were between 5 months and 6 years of age, with a mean age of 3.25 years. The aims of the trial were fully explained to the parents of the children before consent was obtained for their inclusion.

Prior to the start of the feeding period, all the children received Mebendazole (Dosage as recommended for weight and age of each child by the Universidad Sanitaria, Oruro.) to remove any intestinal parasites that may have been present. They were monitored for reappearance of parasites but no further treatment proved necessary throughout the trial period.

Supplements and their Distribution.

The children were randomised into three groups; A, B or C. Each group received a different supplement. The supplements were formulated to contain approximately 175 kcalories and 4.25g of protein in a 200ml drink flavoured with banana. Supplement A was a milkbased supplement, supplement B was based on LC and supplement C contained both LC and 25mg ascorbic acid; as shown below:

The Approximate Composition of the Supplements.

Dried Milk (g) 14.9 NONE NONE
Dried LC (g) NONE 6.0 6.0
Sugar (g) 4.5 18.6 18.6
Raw Banana (g) 55.0 55.0 55.0
Ascorbic Acid (mg) NONE NONE 25.0
Water (g) 126.0 120.0 120.0


The ingredients for each supplement were weighed out weekly and stored in opaque plastic bags to prevent destruction of vitamins during storage. Each groups' supplement was made up in bulk daily and a measured 200ml given to each child. Local women were trained in the preparation and distribution of the drinks, which were repeatedly monitored to ensure consistency of composition.

Supplement A had a negligible iron content, whereas supplements B and C contained approximately 6.33 mg of iron, supplied by LC:

The Approximate Nutritional Composition of the Supplements
(calculated from Division National de Nutricion 1981 data):

Energy (kcals) 175.20 175.40 175.40
Protein (g) 4.25 4.25 4.25
Iron (mg) 0.45 6.33 6.33
Vitamin A (RE) 28.00 902.60 902.60
Vitamin C (mg) 5.00 5.12 30.12

Typically, each child received the appropriate supplement for 6 days a week before breakfast, over a 20 week period. Because of public and religious holidays the actual number of feeding days was 112 out of a possible 140. Attendance, compliance and any adverse effects arising from the supplementation were noted.


The children attended the local health centre on the same day of every month, where height and weight were measured by trained personnel. Information about the incidence and duration of illness (diarrhoea, respiratory infections, nausea or vomiting and fever) was obtained from the mothers of the children retrospectively.

At the initiation and termination of the trial blood haemoglobin was assessed. Blood was taken from the children by a sterile fingerprick technique and the amount of haemoglobin present determined immediately, using the Sahli method. All haemoglobin assessments were done 'blind', that is the person responsible for analysing the blood samples was not aware from which child they had been taken.

Fifteen of the initially recruited sample did not enter the trial because of a local belief that the taking of blood results in death. Their mothers did, however, give permission for the taking of monthly height and weight measurement recordings and this group therefore acted as a control for these parameters.


39 of the 45 children completed the 20 week feeding period. An average attendance of 95% was maintained throughout the project, it did not alter significantly between groups. All the supplements were accepted by the children and no adverse reactions to them were noted.

The monthly height measurements revealed that all groups, including the control, gained in height but at a rate less than the WHO standard for their age. There was no significant difference in the change in height between the groups:

Average Initial and Final Height (Ht) Gains, Compared with World Health Organisation Standards for Age and Time Period (WHO S/A).
Mean Age (months) 28.0 36.0 38.0 45.0
Initial Ht (cms) 77.8 83.5 82.4 88.9
Final Ht (cms) 80.6 86.2 85.3 91.5
Gain in Ht (cms) 2.8 2.7 2.9 2.6
WHO S/A (cms) 3.0 3.0 3.1 2.8
Ht gain as % of WHO S/A 93.0 90.0 95.0 93.0

The weight measurements showed that those children receiving supplements all displayed significant catch up weight gains, that is at a greater rate than the WHO standard for age, whereas the control group did not (p<0.01). There was also a highly significant weight gain for age in group B over groups A and C (p<0.01). Although group C had a greater weight gain for age than group A, this was not significant (p<0.05):

Average Initial and Final Weight (Wt) Gains, Compared with World Health Organisation Standards for Age and Time Period (WHO S/A):
Mean Age (months) 28.00 36.00 38.00 45.00
Initial Wt (kg) 9.85 11.50 11.10 12.70
Final Wt (kg) 10.50 12.40 12.40 13.80
Gain in Wt (kg) 0.65 0.90 1.30 1.10
WHO S/A (kg) 0.90 0.75 0.75 0.70
Wt gain as % of WHO S/A 72.00 120.00 173.00 157.00

It was noted that all the supplements reduced the incidence of morbidity, although there was no difference in the improvement between groups. A rapid improvement in apparent resistance to infection was noted during the first month of the trial; after this time a more constant level was reached. Unfortunately the mothers of the control group children were not willing to give morbidity information so the figures do not show the effect of feeding on morbidity, nor do they allow for seasonal effects.

Monthly Total Morbidity in Days (Includes incidence of Diarrhoea, Nausea, Vomiting, Respiratory Infection and Fever).
Initial Illness per month. 11.5 12.3 16.2
Final Illness per month. 4.4 3.2 3.1
Decrease in Morbidity. 7.1 9.1 12.1
% Decrease in Morbidity. 62.0 74.0 81.0

There was no significant difference in haemoglobin concentration between the groups at the initiation of the study. Group A showed no significant increase in haemoglobin during the trial period. The increase in group B from 8.45 to 9.87 g/dl proved significant using a paired ttest (p<0.05). The increase in group C from 8.88 to 11 .83 g/dl, was greater than the increase shown by group B and also proved more significant (p<0.01).

Blood Haemoalobin (Hb) concentration at initiation and Termination of the trial:
Group:   Initial   Initial   Final Hb   Final
  Hb (g/dl): number per group with Hb less than 11g/dl: (g/dl): number per group with Hb less than 11g/dl:
A 8.94 10 (of 12) 9.04 10 (of 12)
B 8.45 11 (of 12) 9.87 9 (of 12)
C 8.88 14 (of 15) 11.83 5 (of 15)


The cultural opposition to blood sampling resulted in the project being refused by the mothers of 15 of the originally recruited children. The other children recruited at random to replace them were all willing to give a blood sample. However, a study becomes less representative if a third of its initial sample withdraw from it. It was noted that those children who refused the blood sample, and who had thus acted as the control, had a mean age of 28 months whereas those children included in the trial had a mean age of 39.6. Indeed, the cultural reluctance to give blood samples was seen to be stronger amongst the more vulnerable groups in Bolivia such as pregnant or lactating mothers and young children.

The supplementation of the children had no significant effect on their height, they continued to grow at the same rate as the control. Changes in height over a short period of time are often difficult to monitor (WHO, 1983) and so these results are not unusual. However, the children receiving the supplements exhibited significant gains in weight compared to the WHO expected gain for their age during the trial period.

175 kcal represents almost 14% of the Bolivian Recommended Daily Intake for energy and 4.25g of protein is approximately 18% of the average requirement for protein for the healthy 6 month to 6 year age group (Division National de Nutricion, 1981). The weight gains of those supplemented, compared with the control group, was therefore not unexpected. The decrease in the incidence of illness during the trial may account for some of this weight gain although the decrease in illness may itself be a result of food supplementation. The lack of information about the morbidity of the control group means that the effect of the supplementation on illness or weight gain cannot be defined. The fact that all groups showed a similar decrease in morbidity suggests that the LC supplement had no demonstrable effect over the milkbased one on the incidence of illness in this trial. Both the control and supplemented groups received Mebedazole, so any weight gain resulting from the elimination of intestinal parasites would be constant between groups.

Why group B should have gained in weight at a significantly higher rate than groups A or C is unclear. All groups were isocaloric and isonitrogenous, although group A got its protein from a milk source whereas the other group did so from the leaf concentrate. Group C also contained an additional 25mg vitamin C. It is possible the higher initial incidence of illness within group C had an effect on the rate of weight gain. If this was the case it may be that LC has a more beneficial effect on weight gain than milk powder, which would support the work of Shah et al (1981).

Iron supplementation, either alone or in conjunction with other nutrients, is necessary to treat iron deficiency anaemia. A therapeutic dose is normally in the order of 3mg iron per kg body weight per day, the equivalent of approximately 34.5mg daily for the children involved in this trial. This amount should reverse anaemia within two months in the majority of cases (Dallman, 1977). As the iron given in this trial was not in a therapeutic dose a rise in blood haemoglobin to a normal level would be expected to take longer, although it has been shown that the absorption of nonhaem iron, in those with a poor iron status, is far greater than in those having a normal status (FAO/WHO 1988). Due to the cultural reluctance for blood sampling, it was unfortunately impossible to attempt a series of blood tests to assess when the supplementation of iron was most effective in this trial. It was shown, however, that the giving of nonhaem iron to these children increased their haemoglobin levels and that giving vitamin C concurrently enhances this effect. The fact that nonhaem iron is best absorbed when it is in its more soluble reduced state and reducing agents, such as ascorbic acid, favour the reduction of iron (Chanarin, 1988); explains this phenomenon.

Although the values for haemoglobin were obtained using a "blind" analysis technique, the equipment could not be definitively calibrated due to the lack of specialised equipment available. As such, values should be treated as relative and not absolute. Further work should involve the use of more sophisticated iron deficiency assessment techniques, using other parameters as well as haemoglobin.

The actual nutritional composition of LC used for this survey was calculated from local samples independently analysed in Mexico, the USA and Sweden. Recent analyses in Bolivia (Ahman, personal communication, 1992) suggest that the iron content of Bolivian produced LC could be up to 50% greater than previous analyses showed (it has been suggested that the higher iron content of the Bolivian samples may have resulted from contamination by the newly installed LC processing machinery). The results of this project must be considered in the light of this as the participating children may have been receiving up to 9.36mg daily, although they were assumed to be receiving only 6.24mg. Such variation in LC composition is important when using LC as an iron source in supplementation programmes and could cause it to be rejected from use when a doseresponse relationship is more critical.

In children aged six months or over, a haemoglobin concentration above 11g/dl is widely accepted as normal (Woodcock, 1987). There is very little aged related change in values between 1.5 and 4.5 years (Hinchcliffe, 1987). However, it has been suggested that "normal" levels of haemoglobin at high altitudes are somewhat greater than those necessary for the rest of the population. A 4% rise in haemoglobin concentration per 1000 metres above sea level, as suggested by Cartwright (1979), would give a "normal" value of approximately 12.7 g/dl. If this is the case, the amount of LC or the duration of this study was possibly not sufficient to meet the increased iron requirement of children on the Bolivian Altiplano, although the issue of altitude and haemoglobin standards is yet to be well defined.


It is clear from this study that many questions remain unanswered. The bioavailability of the iron present in LC has yet to be determined and the practicalities of using it to either prevent or treat anaemia have yet to be addressed.

Issues such as the cost of LC relative to other iron sources, how much can or should be given to malnourished children and whether it would be more effective if used to prevent, rather than treat, anaemia are yet to be resolved.

Leaf Concentrate was revealed in this trial to be an effective means of raising blood haemoglobin levels; its efficacy being enhanced by the concurrent ingestion of vitamin C.


The Universidad Sanitaria, Oruro; R. Griggs & Co., Luis Fuentes, Prudencia Lopez, Rosa & the mothers and children of Machacamarca. This work was funded by Find Your Feet and the Carol Martin Memorial Award.

APPENDIX 1: The Approximate Composition of Leaf Concentrate (LC).

Protein (g) 60.31
Energy (Kcals) 343.43
Calcium (mg) 191.48
Iron (mg) 96.30
Vitamin A (mgRE) 14.57

*These value are a mean of 3 local samples of dry LC independently analysed in the USA, Sweden and Mexico, and were used to I formulate the ' supplements. A sample subsequently analysed by the University of La Paz, Bolivia, found similar values for all nutrients except for iron. The La Paz results suggested that the value for iron was in fact 50% greater than the results given by the analyses carried out in other countries.


Cartwright, G.E. (1979): Diagnostic Laboratory Haematology.

In: Iron Deficiency in Infancy and Childhood; A Report of the International Nutritional Anaemic Consultative Group. pp30. INACG/WHO, Geneva.

Chanarin, I. (1988) Anaemias and Coagulation Disorders of Nutritional; Origin. In: Nutrition in the Clinical Management of Disease (2nd ed.) pp281 Eds. Dickerson, J.W.T. and Lee, H.A. Hodder & Stoughton Ltd, London.

Chwang, L; Soemantri, A. G; Pollitt, E. (1988): Iron Supplementation and Growth of Rural Indonesian Children. American Journal of Human Nutrition, 47 , 496501.

Dallman, P.R. (1977) Nutritional Anaemias. In: Pediatrics (16th ed.) pp l1-19. Ed. Rudolph, A. AppletonCenturyCrofts, NY.

Division National de Nutricion (1981): Recomendacion Diaria de Calorias y Nutrientes para la Poblacion Boliviana. Ministerio de Prevision Social y Salud Publica, La Paz, Bolivia.

FAO/WHO (1988): Requirements of Vitamin A, Folate, Iron and B12. FAO, Rome 1988. Report number 23.

Hinchcliffe, R.F. (1987): References Rages and Normal Findings. In: Practical Paediatric Haematology. Eds. Hinchcliffe, R.F. and Lilleyman, J.S. John Wiley and Sons.

Olatunbosun, D. A; Adadevoh, B. K & Oke, 0. L. (1972): Leaf Protein: A New Protein Source for the Management of Protein Calorie Malnutrition in Nigeria. Nigerian Medical Journal 2 (4) 195-199.

Pollitt, E; Saco-Pollitt, C; Leibel, R.1 and Viteri, F.E (1986): Iron Deficiency and Behavioral Development in Infants and Preschool Children.

American Journal of Human Nutrition, 43 , 555-565.

Shah, F. H; Salam Sheikh, A; Farrukh, N & Rasool, A. (1981): A Comparison of Leaf Concentrate Fortified Dishes and Milk as a Supplement for Children with Nutritionally Inadequate Diets. Qual. Plant. Plant Foods & Human Nutrition, 30 , 245.

UNICEF (1989 - 1994): Para La Vida. Programa de Cooperacion del Gobierno de Bolivia y UNICEF. UNICEF Publications.

Waterlow, J.C. (1962): The Absorption and Retention of Nitrogen from Leaf Protein by Infants Recovering from Malnutrition. British Journal of Nutrition, 16 , 531-540.

WHO (1983): Measuring Change in Nutritional Status. WHO, Geneva.

Woodcock, B.E. (1987): Iron Deficiency and Iron Overload. In: Practical Paediatric Haematology. Eds. Hinchcliffe, R.F. and Lilleyman, J.S. John Wiley and Sons.


About LFL    Library    FAQ's    5-a-Day    Leaf Crops    Leaf Concentrate    Español    Contact Us    You Can Help

[ Home Page ]  Copyright © by Leaf for Life    [ Email: Information from Leaf for Life ]