Socioeconomic disadvantage and risk of low birth weight

The association between socioeconomic disadvantage and low birth weight is well known. We explored the impact of disadvantage on birth weight by undertaking spatial analysis of all births occurring to residents of New South Wales, Australia, from 1990-2003. We also investigated the influence of other covariates including smoking and remoteness. We geocoded the place of residence for 1,210,858 of 1,221,079 births, and assigned each to one of 587 postcodes. We determined the socioeconomic status of each postcode using the Australian Index of Relative Social Disadvantage, a composite Census derived index that combines variables relating to education, occupation, non-English speaking background, Indigenous origin and the economic resources of households. We examined and mapped the spatial distribution of this data, using Conditional Autoregressive (CAR) modeling to determine the relationship between the relative disadvantage of each area and low birth weight (<2.5 kg). Mean birth weight was 3.37 kg with a standard deviation of 600g. 93,723 births were of 2.5 kg or less, and CAR analysis suggested that 56% of the variation in risk of low birth weight between areas was spatial. An interquartile range increase in disadvantage was associated with a 15% increase in the risk of low birth weight, 7% increase in the risk of a pre term birth and 20% increase in the risk of a small for gestational age baby. The provision of maternity and antenatal care needs to take account of the increased birth risks associated with neighborhood disadvantage.