Our study evaluated several outcomes in the first four years of age in three population-based birth cohorts in Southern Brazil. Although children in each cohort were born 11 years apart, we performed pooled analyses of the three cohorts because statistical tests did not show any evidence of interaction between each cohort and gestational age groups.
One limitation of this study, which is common in birth cohorts, was attrition rates in the follow-up visits. However, with the exception of the 12 months visit of the 1982, in which we failed to trace 20.7% of the children, we were able to locate at least 85% of the children in all other visits. Losses in 1982 were more frequent among the poorest and the richest strata of the population, as middle-class families were more easily found [17
]. Also, as the estimation of gestational age was based on the date of the last menstrual period, we had 15.6% of missing cases.
We found that preterm babies, even those born between 34–36 weeks of gestation, children born with 37 weeks of gestation, and those born with 42 or more weeks were at increased risk of death in the first month of life, relative to children born between 39–41 weeks. Even after adjusting for possible confounders, compared to the reference groups of 39–41 weeks, the relative risks of neonatal death were more than 30 times higher for babies born before 34 weeks, 3.4 times higher for those born between 34–36 weeks, nearly three time higher for babies born at 37 weeks, and the double for post-term children. Preterm children and those born with 37 weeks of gestation were also at increased risk of infant mortality.
In addition to increased mortality risks, we found that all preterm children - including those born between 34–36 weeks - were more likely to present with stunting and underweight at 12 months of age, but this was no longer observed at 48 months of age. We also found that children born at 37 weeks were more likely to present underweight at 12 months of life. Post-term children presented increased risk of death in the first month of life.
Our finding of an increased risk of mortality and deficits in growth not only babies currently defined as preterm (<37 weeks), but also amongst those born at 37 weeks is a cause for concern. In Pelotas, where the whole gestational age curve has shifted to the left in the last three decades, and the prevalence of preterm births increased from 6.2% in 1982 to more than 15% in 2004 [21
]. Births with 37 weeks also increased from 7.1% in 1982 to 11.4% in 2004. The causes for this increase in preterm births are not clearly elicited yet, but it is possible that interruption of pregnancies may have played an important role, as the proportion of cesarean sections doubled between 1982 and 2004 [21
Finally, it is not possible to ascertain in an observational study if the increased risks observed among infants born before full maturity were due to early exposure to the extra-uterine environment, or damages produced by maternal conditions that may also have produced the untimely birth, or even a combination of both situations. A previous study suggests that preterm birth and exposure to maternal medical conditions are independent risk factors for neonatal morbidity, with the former playing a stronger role [11