Prevalence of Child Undernutrition in India: Estimating Extent of Deficits Using Distributions of Nutritional Outcomes

Anthropometric indicators height for age, weight for age, and weight for height, calculated from the data of the children’s height, weight, and age are transformed to Z-scores in order to assess nutritional status of under-five children. Following World Health Organization (WHO) classification scheme, the degree of prevalence in a population is defined as the percentage of subjects whose Z-scores are lying more than two standard deviations below the reference median. Although this formulation facilitates a gross comparison among sub-populations under question, two of its shortcomings are noteworthy. First, it fails to capture the salient distributional aspects of sample Z-scores different from those of the reference population and thereby the nutritional status estimates may be significantly biased. Second, empirical evidences suggest that the distribution under question can be skewed and substantially away from normal for the developing countries. The present article shows that for the selected states of India, distribution of Z-scores follows a general class of skew normal distribution, in which the reference population is a member and compares two and more members in this family. The degree of nutritional deficit is then quantified in terms of the population parameters. Empirical illustrations are given using data on Z-scores for selected Indian states (IIPS and Macro International 2007). The findings are indicative of the existence of comprehensive gaps in the perceived level of undernutrition prevalence for the selected Indian states.