Anaemia is a major contributor to the global disease burden and half of pregnant women in India were anaemic in 2016. The aetiology of anaemia is complex, yet anaemia determinants are frequently examined in isolation. We sought to explore how shifts in sociodemographic (wealth, age at pregnancy, education, open defecation, cooking fuel type, household size), programmatic (iron–folic acid tablet consumption, antenatal care visits) and dietary factors (intake of Fe, folic acid, vitamin B12, phytate) predicted changes in anaemia prevalence.

Nutrient levels for eighty-eight food items were multiplied by household consumption of these foods to estimate household-level nutrient supply. A synthetic panel data set was created from two rounds of the District Level Household and Facility Survey (2002–04 and 2012–13) and Household Consumer Expenditures Survey (2004–05 and 2011–12). Ordinary least-squares multivariate regression models were used.

Districts (n 446) spanning north, north-east, central and south India.

Pregnant women aged 15–49 years (n 17 138).

In the model accounting for both non-dietary and dietary factors, increased age at pregnancy (P<0·001), reduced village-level open defecation (P=0·001), consuming more Fe (P<0·001) and folic acid (P=0·018) and less phytate (P=0·002), and urbanization (P=0·015) were associated with anaemia reductions. A 10 mg increase in daily household Fe supply from 2012 levels was associated with a 10 % reduction in anaemia.

Public health interventions to combat anaemia in pregnant women should use a holistic approach, including promotion of delayed marriage, construction and use of toilets, and measures that facilitate adoption of nutrient-rich diets.