BMC Pregnancy Childbirth 2014 May 1;14:155. Epub 2014 May 1.
Departamento de Puericultura e Pediatria, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo, Hospital das Clínicas de Ribeirão Preto, 7°, andar, Av, Bandeirantes, 3900, Ribeirão Preto, SP 14049-900, Brazil.
Background: To analyze trends in LBW (low birth weight) rate using birth registry data and identify factors associated with LBW in São Luís comparing two birth cohorts separated by a 12-year interval.
Methods: 2,426 births were included in 1997/98 and 5,040 in 2010. The dependent variable was LBW (<2,500 g). Multiple logistic regression was performed to determine the association of independent variables with LBW. Data were also obtained from SINASC (Brazilian National Birth Registry) to analyze stillbirth and LBW rates trends from 1996 to 2010, using 3-year moving averages.
Results: LBW, intrauterine growth restriction (IUGR) and preterm birth rates did not differ between the two cohorts. Despite this, birth registry data showed increasing LBW rate up to 2001, coinciding with decreasing stillbirth rate. Both stillbirth and LBW rates decreased thereafter. A significant reduction was observed in the percentage of teenage mothers, mothers with up to 4 years of education, family income up to one minimum wage and mothers who did not attend prenatal care. There was an increase in maternal age ≥35 years and schooling ≥12 years. The variables associated with LBW in 1997/98 were young maternal age (<18 years), maternal smoking during pregnancy and primiparity. Variables that remained in the adjusted model in 2010 were female gender, income <3 minimum wages, lack of prenatal care, maternal smoking during pregnancy and primiparity.
Conclusions: Although LBW rate did not differ between the two cohorts, this apparent stability masked an increase up to 2001 and a decrease thereafter. The rise in LBW rate paralleled reduction in the stillbirth rate, suggesting improvement in obstetrical and newborn care. Maternal, socioeconomic and demographic factors associated with LBW differed between the two cohorts, except for smoking during pregnancy and parity that were significantly associated with LBW in both cohorts.