Contents - Previous - Next

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

Analysis of complementary feeding programme participation: A milk-cereal mix for preschoolers

Eugenia Muchnik and Isabel Vial,
with the collaboration of
Teresa Boj, Paula Galvez. and Gloria Cardenas

In 1985 social expenditures in Chile amounted to an estimated US$2.4 billion, which was close to 15% of the gross domestic product. Approximately 3% of this amount was allocated to supplementary feeding programmes (SFPs), which have been a traditional component of the social expenditure agenda of the public sector [1]. Together with other important programmes such as health, education, housing, and family monetary subsidies, SFPs contribute to the redistribution of income to poorer households.

The SFPs have been modified several times to target better the allocation of benefits to low-income households and thus improve efficiency and equity. In fact, health and nutrition indicators in Chile do show continuous improvement in spite of the two economic recessions that have taken place since 1975. A study was conducted in Santiago to assess poor households' behaviour in the demand for milk-cereal, a supplementary food for preschoolers introduced in 1975 as part of the main SFP, namely the complementary feeding programme (CFP).


The study


An important modification of the CFP introduced in 1975 was the substitution of a milk-cereal mix for powdered milk for the 1-5-year-old beneficiaries [2]. The purposes of this change were to decrease the cost of the programme by using relatively cheaper raw materials, and to target the low-income population more effectively. It was expected that leakages to adults within the household would diminish, given that cereal mixtures were not a typical component of adult diets.

Questions on the quality and acceptability of the milk-cereal mix delivered by the programme had been raised earlier and apparently were resolved adequately. Some concern remained as to actual acceptability, however, since the product was never made commercially available, as had been the case with the powdered milk delivered by the programme. Also, participation of potential beneficiaries in the milk-cereal programme was lower in 1986 than participation in the milk-distribution programme.

To provide adequate answers to these questions, the present study was undertaken in 1987 with the following objectives: (1) to assess the degree of acceptability of the milk-cereal product delivered by the CFP to preschoolers attending the National Health System (NHS) in Santiago; (2) to identify non-participants within the NHS clientele, particularly in terms of socio-economic household characteristics, nutrition status, and biomedical risk of the children; (3) to assess the net increase in calorie consumption of children participating in the CFP; and (4) to estimate the average monthly income transfer represented by the milk-cereal distribution.

The main purpose was to provide the necessary information for revising the programme in at least two aspects: the quality of the product delivered to preschoolers, and the coverage of these children, especially those from low-income households. It was thought that better targeting was required in both respects.


The sample

Twenty of the 66 NHS health centres in urban Santiago were randomly selected, representing all six health subsystems into which metropolitan Santiago is divided. In a second stage, the preschool children were stratified into two age-groups, 12-23 months and 24-72 months old, with a sample size proportional to the population in each health centre. The final sample, after data cleaning, comprised 215 children 1 year old and 785 children 2-5 years old, which represents close to 1% of the population. Each household was surveyed and the mother interviewed at home between August and November 1987. Information was gathered on socio-economic household characteristics, participation in the CFP and acceptability of the food distributed, physical and economic accessibility of the health centres, and the target children's recent histories of illness and mothers' subsequent behaviour in seeking health services. In addition, the height and weight of the children were assessed using the salter scale and vertical boards. Data on weight and height at birth and for each health examination during the reference period were obtained from the health centre records.

TABLE 1. Distribution of the sample and subsample by income decile

Decile Upper income limita Total sample Subsampleb
N % N %
1 2,095 118 15.9 60 14.3
2 3,072 126 16.9 74 17.6
3 4,189 125 16.8 59 14.0
4 5,236 96 12.9 65 15.5
5 6,649 77 10.4 50 11.9
6 8,406 93 12.5 58 13.8
7 11,171 61 8.7 32 7.6
8 15,639 27 3.6 14 3.3
9 25,832 16 2.2 4 1.0
10 >25,833 5 0.6 4 1.0
Total   744 100.0 420 100.0

a. Monthly household income per capita in November 1987 pesos [5].
b. Subsample from which food-consumption information was obtained by 24-hour recall.

To complement this information, a subsample of households was randomly chosen for a food-consumption survey, consisting of two 24-hour recalls by all household members; these included a weekday and a weekend day, to account for differences in patterns. The weights given to each of these were 5/7 and 217 respectively. The final size of the subsample was 420 households.

The distribution of the sample and subsample by income decile is shown in table 1. The income deciles were defined according to the per capita household income limits derived from the National Socioeconomic Survey [4]. It is worth noting that 62% of the sample belonged to the four lowest deciles.

Contents - Previous - Next