Objective: Data from two household surveys on infant and child health status undertaken in the mid-80s and mid-90s, complemented with previous data collected from maternity hospitals records and more recent data provided by the state system on birth registries, allowed to characterize and analyse secular trends in birth weight in the city of S. Paulo, Brazil.
Methods: The household surveys included random samples of children under 5 years old (n = 1,016 children in 1984-85 and n = 1,280 children in 1995-96). A random sample of births that took place in the city's hospitals in the year of 1976 (n = 5,734) was drawn from the hospital records. Birth registries refer to children born in the city between 1993 and 1998 (around 200,000 per year). The study of the social distribution of birth weight took into account the per capita family income and maternal schooling. For the analysis of the determinants of secular trends, hierarchical causal models, multivariate regression analyses and calculations analogous to the ones used to assess population attributable risks were applied.
Results/conclusions: Birth weight distribution in S. Paulo city (an average of 3,160 g and 8.9% of the values < 2,500 g) is below the expected for optimum fetal growth conditions (average 3,400-3,500 g and 4-5% of the values < 2,500 g). The birth weight distribution did not change substantially along the study period (1976-1998). However, there are evidences of changes when different socioeconomic strata are considered separately. Among the lower strata trends have been positive and this seems to be due to increases in intrauterine growth as a result of an improvement in family's purchasing power, women's weight and height, prenatal care and, possibly, the reduction in smoking. Among the higher socioeconomic strata, birth weight trends have been negative apparently due to an increase in premature births of unknown origin.