Hierarchy of Associations Between BMI-for-Agez-Scores, Growth and Family Social Status Among Urban Bengali Girls of Siliguri Town, West Bengal: A St. Nicolas House Analysis

Abstract

The present study assesses the association of different socio-economic variables with children’s BMI-for-age z-scores (BAZ) and influence of children’s height on BAZ, a proxy index of nutritional status. The study was undertaken among 322 girls belonging to the Bengali Hindu Caste Population (BHCP) aged 5–13 years. These girls were the students of two schools located in Siliguri town of West Bengal, India. The prevalence of short stature, underweight and overweight was assessed using the World Health Organization’s ( WHO, 2007 , Growth Reference Data for 5–19 years) references. The statistical analyses used were St. Nicolas House Analysis (SNHA), analysis of variance (ANOVA), linear and logistic regression. The prevalence of underweight, overweight and obesity was 19.3%, 17.8% and 5.9%, respectively. The prevalence of short stature (−2.0 HAZ) was 8.7%. The range of BAZ was from −5.69 to 4.15. The variation of BAZ explained by height-for-age z-scores (HAZ) was 11.4%. The BAZ was observed to be associated with mothers’ occupation as revealed by 2% variation through SNHA analysis. The present study observed the usefulness of SNHA for non-parametric data with unequal sub-sample or categories. However, SNHA was not devised to assess the direction and magnitude of variables of interest. The finding of the present study supports the use of BAZ as a proxy adiposity measure among the overweight/obese populations and populations with normal growth in height. The study further supports the recommendations that mother empowerment can help improve nutritional status of a girl child.

In Python available: St. Nicolas House Algorithm (SNHA) with bootstrap support for improved performance in dense networks

“…The SNHA is a robust, non-linear and parameter free visualization method for multivariate data. It was developed recently and verified in a few studies (Groth et al 2019;Dorjee et al 2021;Hermanussen et al 2021;Scheffler et al 2021). It displays associations between variables in a graph yielding immediate insights into the principle data structure.…”