Faculty of Public Health - Andalas University - OCS, 13th IEA SEA Meeting and ICPH - SDev

Font Size: 
Additive Quantile regression to model undernutrition among under-five children in Uttar Pradesh, India

Last modified: 2018-08-07


Background: Nutrition-related factors are responsible for about 35% of child deaths and 11% of total global disease burden. In Uttar Pradesh, the most populous state of India, about half of the children are stunted (low height-for-age), 40% are underweight (low weight-for-age) & wasting (low weight-for-height), being a serious problem is affecting 18% of children under five.


Materials and Methods: Unit-level data on 36,465 children aged 0-59 months taken from the National Family Health Survey (NFHS- 4) ,2015-16 has been used for the study. Statistical analyses were performed using STATA 14.0 and R 3.5.1 software. Additive quantile regression model (QRM) was fitted to model linear & non-linear effects of significant covariates on entire conditional distribution of response variables. Hundred bootstrap samples were generated for model estimation to obtain 95% bootstrap confidence intervals of the estimate. Optimum number of iterations was determined by cross validation.


Results: Quantiles for response variables were fixed according to prevalence. Child’s age showing largest absolute effect size, emerged as the most significant non-modifiable factor affecting undernutrition.


Conclusions: On comparing results with traditional logistic regression model, it was found for variables like undernutrition where extreme quantities are of more interest than analysis of means, QRM may be preferred.


Keywords: Undernutrition, Children, Additive Quantile Regression