Studies concerning very preterm newborns are either defined by birth weight (<1500 g) or gestational age (<32 weeks). The aim of our study was to underline limits of cohort definitions by birth weight.
Methods: Data come from the Nord Pas de Calais EPIPAGE cohort. Every birth occurring in 1997 before 32 weeks or with a birth weight less than 1500 g and transferred in a neonatal unit was included. Two cohorts were defined, one by gestational age (<32 weeks), the other by birth weight (<1500 g). Two subgroups could be defined from these to cohorts: group A (<32 weeks and > or =1500 g), from cohort (<32 weeks), group B (> or =32 weeks and <1500 g) from cohort (<1500 g).
Results: Five hundred nine newborns were included. Perinatal characteristics of both cohorts seemed comparable. The analysis by subgroups A and B revealed an excess of pulmonary and neurological morbidity in very preterm infant compared to very low birth weight newborn. This was linked to an excess of growth restricted newborns in this cohort with more advanced gestational ages.
Conclusion: Cohorts of very preterm newborns should rather be defined by gestational age. If not possible, results in very low birth weight cohorts should also be given by gestational age and rate of growth restriction should be described.