Is Groundwater Contamination Contributing to Women's Anemia in India? Using Machine Learning Approach for Risk Zone Predictions

Sourav Biswas, International Institute for Population Sciences (IIPS)
Aparajita Chattopadhyay, International Institute for Population Sciences, India
Aditi Prasad, International Institute for Population Sciences (IIPS)

Groundwater contamination poses a serious concern for Indian women, given that a significant portion rely on this source for their drinking water. This study aims to explore the understudied link between water pollutants and anemia by analysing groundwater contaminants and anemia. The study also contributes to a critical knowledge gap by predicting anemia risk zones. We used the integrated data of 581,286 women from NFHS-5 (2019-21) and the groundwater contaminants from CGWB (2019-21). Multivariate regression and spatial analysis were used to study the linkages at the individual and district levels, while machine learning (RF, ANN, XGBoost) techniques were employed to predict anemia risk zones. The findings reveal that high exposure to contaminants like arsenic (above 0.01 mg/L), nitrate (above 45 mg/L), fluoride (above 1.0 mg/L), sulphate (above 200 mg/L), and pH (above 8.5) have 10%, 7%, 3%, 6% and 32% more likely to be anemic as compared to there above permissible limits. LISA cluster map indicates high arsenic contamination and anemia prevalence are notable in Eastern and North-Eastern India. The study further predicts high anemia risk zones in Eastern, North-Eastern, Western, and South-Central regions, underscoring the critical need for interventions to improve groundwater quality and address anaemia nationwide.

Keywords: Population, Environment, and Climate Change, Health and Morbidity, Gender Dynamics, Population Policies

See extended abstract.