Studies of outcomes of late preterm (LP) birth, defined as delivery at 32 to 36 6/7 weeks’ gestational age (GA), have implications for the largest and fastest growing segment of the preterm population. Neonatal mortality and morbidity are more common in LP children than in those born full term (FT) with normal birthweight. A recent survey indicated poorer developmental and learning outcomes in young LP children compared with FT controls.

Lipkind et al added to this literature by documenting the negative impact of LP birth on academic achievement in a sample of school-age children born in New York City and attending the New York City public schools at the time of the outcome assessments. By examining the effects of GA on outcomes controlling for sociodemographic factors, admission to a neonatal intensive care unit, maternal risks, and complications of labor and delivery, the findings help to isolate the deleterious effects of shortened GA from those related to other factors.

The linear relation of increasing GA from 32 to 39 weeks with more positive outcomes also suggests that any shortening of the normal gestational period may carry some risk for adverse effects on children’s learning abilities. This finding is consistent with another large-scale study conducted in Scotland as well as with those of prior studies.

As acknowledged by the authors, one of the study’s major limitations is that the sample was not representative of the regional population of LP births. The high rates of Medicaid assistance and parents with less than a high school education suggest that the sample overrepresented families at relative social disadvantage. If so, and given evidence that the negative effects of LP birth on developmental outcomes are amplified by childhood poverty, differences between the LP and FT groups in this study may be more pronounced than would be anticipated in a more representative sample. In addition, neonatal medical complications that occurred in infants not admitted to a neonatal intensive care unit or that precipitated hospital readmissions were not taken into account in assessing the effects of shortened GA.

Another limitation is that the authors did not examine group differences in rates of impaired scores on tests of reading and mathematics but only in continuous measures of these abilities. This makes it difficult to gauge the clinical significance of the LP group’s achievement scores because we are unable to determine the relative risk of deficits in academic achievement in comparison with FT children.

A related concern is that the use of curriculum-based assessments of achievement provide useful measures of how the children were responding to instruction but tell us little about children’s learning abilities relative to broader normative standards, as would be measured using nationally standardized achievement tests. Although group differences in the rates of special education (25% in the LP group vs 19% in the FT group) suggest a modest increase in the need for this assistance in the LP group, it is also difficult to gauge the clinical significance of this difference. The achievement scores are not linked in time to the special education placements, and the basis and uniformity of these placements are unknown.

Future research is needed to replicate the finding of adverse educational consequences attributable to 32-36 weeks’ GA. In undertaking these studies, it will be important to recruit more representative samples and to take into account residual confounding of GA with factors such as neonatal medical complications. Although beyond the scope of the study by Lipkind et al, it will also be useful to examine the full range of potential predictors of outcome besides GA and to identify factors that make it more likely that LP birth will lead to poor achievement and special education.

Potential predictors in addition to neonatal complications include the following: (1) maternal risks and reasons for early delivery; (2) biomarkers such as indications of inflammation, neural abnormalities, and intrauterine growth retardation; and (3) postnatal medical problems, neurodevelopmental impairments, and environmental influences. Consideration of multiple factors will help to specify which LP children are at highest risk for school-age learning problems. An issue of substantial clinical import will be to determine whether many or most children are adversely affected in some way or whether these negative effects are restricted only to a smaller subset of children.

Longer-term follow-up and inclusion of measures of cognition and behavior are critical in assessing the broader consequences of LP birth on development and in understanding the basis of children’s academic problems. Although prospective studies of LP infants beginning in the perinatal period would ensure uniform and comprehensive assessments of the factors mentioned in the previous text, existing data on labor and delivery, neonatal complications, and early postnatal status can be put to use in identifying predictors of outcomes measured later in childhood.

Other essential elements of sound research designs are the recruitment of FT controls from backgrounds similar to those of LP children, use of measures that are sensitive to the subtle developmental impairments often found in preterm children, and consideration of genetic susceptibilities and experiences at home and school that may buffer or exacerbate the effects of LP birth.

For developmental neuroscientists, investigations of outcomes will shed new light on how perturbations in early brain development related to LP birth eventuate in developmental impairments and how these consequences unfold with age. For the health care providers who work with these children and their families, findings from this research will assist in clinical decision making. More information on perinatal factors related to poor outcomes will increase the precision of estimates of individual risk, and knowledge of outcomes will enhance awareness of how development is affected and thus guide development of improved procedures for early identification and intervention.