Spatio-temporal modeling for malnutrition in tribal population among states of India a Bayesian approach

Abstract

Exploring Bayesian spatio-temporal methods to analyze spatial dependence in malnutrition at the state level for tribal children (less than 3 years) population of India and change over time (three rounds of NFHS-2(1998-99),3(2005-06) and 4(2015-16)). The Bayesian model, fitted by Markov chain Monte Carlo simulation using OpenBUGS, for spatial autocorrelation (through spatial random effects modeling). The model estimated (1) mean time trend and (2) spatial random effects. Results of spatio-temporal modeling for stunting, wasting and underweight exhibited a declining mean trend across the study region from NFHS-2 to NFHS-4. Spatial random effects exhibited spatial dependence for various states in stunting, wasting and underweight tribal children. Future research should analyze spatio-temporal distribution for malnutrition at district level which will require NFHS-5 data. Also, analysis can be done capturing spatio-temporal interaction and identifying hot spots and cold spots at district level.

Keywords: Bayesian spatio-temporal model; Malnutrition trend; Posterior predictive density; Spatial clustering; Spatial random effects; Spatio-temporal.