Do socio-economic inequalities in infant growth in rural India operate through maternal size and birth weight?

Background 3·1 million young children die every year from undernutrition. Greater understanding of associations between socio-economic status (SES) and the biological factors that shape undernutrition are required to target interventions. Aim To establish whether SES inequalities in undernutrition, proxied by infant size at 12 months, operate through maternal and early infant size measures. Participants and Methods The sample comprised 347 Indian infants born in 60 villages in rural Andhra Pradesh 2005-2007. Structural equation path models were applied to decompose the total relationship between SES (standard of living index) and length and weight for age Z-scores (LAZ/ WAZ) at 12 months into direct and indirect (operating through maternal BMI and height, birthweight Z-score and LAZ/WAZ at 6 months) paths. Results SES had a direct positive association with LAZ (Standardized coefficient = 0.08, 95% CI = 0.02, 0.13) and WAZ at age 12 months (Standardized coefficient = 0.08, 95%CI = 0.02, 0.15). It also had additional indirect positive associations through increased maternal height and subsequently increased birthweight and WAZ/LAZ at 6 months, accounting for 35% and 53% of the total effect for WAZ and LAZ respectively. Conclusion Findings support targeting evidence based growth interventions towards infants from the poorest families with the shortest mothers. Increasing SES can improve growth for two generations.