Socioeconomic inequalities in reproductive outcomes in the Italian NINFEA birth cohort and the Piedmont Birth Registry

ABSTRACT BACKGROUND: socioeconomic inequalities in reproductive outcomes have been consistently reported in several countries. In a European collaborative study conducted in 2012 whose aim was to investigate the association between socioeconomic position (SEP), measured through maternal education, and preterm delivery inconsistent results were found for the NINFEA birth cohort. However, NINFEA contributed to that study with the first 2,500 pregnancies only, and estimates were not adjusted for any potential confounders assuming that SEP is a distal exposure, that could not be affected by other preterm risk factors. OBJECTIVES: to investigate the relationship between SEP and the reproductive outcomes using the entire NINFEA cohort and compare the results with the population-based Piedmont Birth Registry (PBR), accounting for potential baseline collider bias both in the cohort and in the registry. DESIGN: observational study. SETTING AND PARTICPANTS: 5,323 NINFEA singletons, whose mothers registered into the study before the 36th week of gestation, were analysed. Analyses on maternal education were replicated in the 2011 PBR of 35,318 singletons live births. Factors affecting the likelihood of being a member of the NINFEA study or becoming pregnant in the general population were treated as potential confounders to adjust for baseline collider bias. MAIN OUTCOME MEASURES: the association of maternal education and a recently developed household income indicator with both preterm delivery (<37th weeks of gestation) and low birth weight (<2,500 gr) were analysed. RESULTS: in the NINFEA cohort, low SEP was positively associated with both preterm delivery and low birth weight, with slightly stronger associations for household income, especially on low birth weight. Results were consistent with those obtained in the PBR data, where an inverse relationship between maternal education and the two reproductive outcomes was found. In both populations, there was confounding due to maternal age and parity, showing that independently of the nature of the source population, baseline factors that affect the probability of being a member of such source population have to be accounted for to allow causal inference. CONCLUSIONS: low SEP is associated with adverse reproductive outcomes in a contemporary Italian population.