Modelling the effects of maternal socio-demographic characteristics on the preterm and term birth weight distributions in Greece using quantile regression.

The present study aims at modelling the effects of maternal socio-demographic characteristics on the birth weight distribution in Greece. The analysis is based on nationwide vital registration micro-data; 103,266 single live births recorded in 2006 are considered. Quantile regression models, allowing for the effects of covariates to vary across the conditional distribution of the dependent variable, birth weight, are applied to preterm and term births separately. The statistical analysis shows that the effects of most factors differentiate across the birth weight distributions. Ordinary Least Squares (OLS) coefficients, on the other hand, systematically underestimate effects at the lower tail and overestimate effects among heavier babies. Hence, quantile regression has a strong advantage over the OLS method. The findings also indicate that birth weight distributions of term and preterm infants are distinct and should be analysed separately. For both distributions female sex, primiparity, age of mother over 35 and prior history of stillbirths and child deaths are related to lower birth weight while higher educational attainment has a protective effect. Among term births, illegitimacy, living in big metropolitan areas and immigrant status of the mother are also significant predictors. For preterm births the impact of age of mother, parity and, in particular, prior stillbirths or deceased children is very pronounced.