INTERACTIONS BETWEEN MALNUTRITION AND DISEASE: A simultaneous equations model applied to data from a low-income area of Karachi

This paper examines the interactions between disease and malnutrition using cross-section data on pre-school children from a low-income area of Karachi. A simultaneous equations model is utilized in the study, the endogenous variables being the number of spells of diarrhoea and respiratory illness reported by the childrens’ mothers during the preceding six months (measure of disease) and three alternative measures of malnutrition. The exogenous variables in the model include parents’ schooling, weekly food expenditures, age and sex of the child, length of breastfeeding and other dietary measures, immunizations, sanitation practices and housing characteristics. The results of the study indicate that there is, in fact, a high positive correlation between the incidence of disease and malnutrition but that, contrary to what is widely believed, the direction of causality appears to run primarily from malnutrition to disease, rather than vice-versa. The study’s findings also indicate that neither the level of family food expenditures nor the education of either parent is significantly related either to the child’s nutritional status or to the frequency of illness. On the other hand, the results point to the possible importance of the mother’s own health and of infant feeding practices as determinants of the child’s nutritional status.