Spatial dependency in child malnutrition across 640 districts in India: need for context-specific planning and interventions.

BACKGROUND

Child malnutrition remains a matter of concern in India as the current levels are high and the decline is slow. National Family Health Survey (NFHS-4, 2015-16) data, for the first time, provides credible, good quality data at district level on social, household and health characteristics.

METHODS

Techniques of spatial analysis on data in respect of 640 districts were used to identify spatial characteristics of the nutrition levels for children in the 0-60-month age group. Further, the principal component analysis (PCA) was used to identify 7 important correlates of the malnutrition out of 21 relevant components provided in the NFHS-4. The paper further uses three techniques, ordinary least squares (OLS), spatial lag model (SLM) and spatial error model (SEM) to assess the strength of correlation between the malnutrition levels and the shortlisted correlates.

RESULTS

The use of SLM and SEM shows improvement in the strength of the association (high R-square) compared to OLS. Women's height and Iodized salt in stunting, child anaemia in wasting, women's height and child anaemia in underweight were found to be significant factors (P < 0.01) along with spatial autoregressive constant.

CONCLUSIONS

Such analysis, in combination with PCA, has shown to be more effective in prioritizing the programme interventions for tackling child malnutrition.