Determinants of Anaemia in Pregnancy in India: A Comparative Weighted Logistic Regression Analysis Using NFHS-4 and NFHS-5

ABSTRACT

Background: In India, the issue of anaemia during pregnancy, defined by low haemoglobin levels, continues as a major public health concern. Recent national surveys indicate a rising anaemia rate among pregnant women, regardless of large-scale nutritional programmes. This study aims to investigate the factors influencing anaemia by comparing data from NFHS-4 (2015–16) and NFHS-5 (2019–21) using a comprehensive analytical method that considers socioeconomic, demographic, nutritional, and environmental factors.

Methodology: Data were obtained from samples that represent the national population of women who are currently pregnant, as recorded in NFHS-4 and NFHS-5. Weighted logistic regression models, adjusted for the complex survey design, were used to estimate adjusted odds ratios (AORs) and 95% confidence intervals. The variables considered were maternal demographics, socioeconomic status, BMI, ANC usage, WASH conditions, and media exposure. The model's performance was checked using AIC, McFadden’s pseudo-R², and ROC curves. Results were considered significant if p was less than 0.05.

Findings: More than half of the people had anaemia in both survey rounds. The characteristics of the population remained largely unchanged. BMI was the most reliable factor in preventing anaemia. Education helped prevent anaemia in NFHS-4 but was less effective in NFHS-5. Wealth became a bit more protective in the later survey. Differences related to caste were more noticeable in NFHS-5.ANC utilisation, WASH conditions, media exposure, and birth interval contributed slightly in adjusted models. Overall model discrimination remained moderate among rounds.

Conclusions: Maternal nutritional status continues to play a crucial role in anaemia risk, while socioeconomic and social-group factors are evolving. To effectively reduce anaemia during pregnancy, it is crucial to enhance nutrition-focused interventions and address structural inequities.

Keywords: Anaemia, Pregnancy, NFHS, Weighted Logistic Regression model.