RESUMO
OBJECTIVES: To compare the classification of preterm postnatal poor growth using healthy preterm vs fetal growth references and to examine associations with neurodevelopmental impairment in infancy and childhood. STUDY DESIGN: We included 613 infants born at <33 weeks of gestation. Using the INTERGROWTH-21st (healthy-preterm growth) reference and the Fenton and Olsen (fetal growth) references, we classified poor growth as a decline in z-score from birth to term-equivalent >0.8 SD (weight), >1 SD (head), and >2 SD (length). We used generalized estimating equations to estimate aOR for neurodevelopmental impairment at 18 months and 7 years of corrected age, comparing infants with and without poor growth by each reference, accounting for multiple births and covariates. RESULTS: The prevalence of poor growth was higher with INTERGROWTH-21st than with fetal references for all measurements. Agreement was higher between the Fenton and Olsen (fetal) growth references (0.72-0.81) than between INTERGROWTH-21st and fetal references (0.41-0.59). Poor growth by fetal references (but not by INTERGROWTH-21st) was associated with low neurodevelopmental scores in infancy and childhood. Poor weight gain using the Fenton reference was associated with 18-month Mental Developmental Index <85 (aOR 1.6, 95%CI: 1.1, 2.4) whereas poor weight gain by the INTERGROWTH-21st reference was not (aOR 1.0, 95%CI: 0.6, 1.7). Poor linear growth by the Olsen reference, but not INTERGROWTH-21st, was associated with 7-year verbal intelligence quotient <70 (aOR 3.5, 95%CI: 1.1, 12.7). CONCLUSIONS: Poor neonatal growth categorized using fetal references showed stronger associations with long term neurodevelopment than poor growth categorized using the INTERGROWTH-21st standards.