P2C333: White Matter Microstructural Correlates of Postnatal Weight Gain in Very Preterm Infants

Background: Infants born very preterm ( < 32 weeks gestational age at birth) account for about 1.6% of live births in the US. Long term neurocognitive delay affects up to half of these very preterm infants (VPIs). Furthermore, while birth weight and post-natal growth have been reported to correlate with neurodevelopmental outcome, effective nutritional strategies to promote growth in this population are not yet standardized, perhaps because the neurobiology and underlying mechanism of these associations remain elusive. Diffusion tensor imaging (DTI) can delineate the microstructural development of neonatal brains and fill in this knowledge gap. Thus, in this study, we utilized diffusion tensor imaging to characterize how birth weight and postnatal weight gain influence VPIs’ white matter (WM) maturation.

Methods: For this retrospective analysis, we included neonates who (1) were born very preterm ( < 32 weeks gestational), (2) had DTI performed as part of their pre-discharge MRI and (3) had adequate data in the electronic health record to obtain weight measurements from birth and the first 4 postnatal weeks. We excluded subjects who had gross structural abnormalities, artifacts, or distortion on their scans and those with neurological deficits noted in their health record during their hospital stay. We then applied both voxel-wise general linear model (GLM) and tract specific statistical analyses to analyze the association of birth weight with WM tract maturation when controlling for gestational age at birth and time between birth and scan. We also conducted similar analyses to analyze the association of weight changes in the first 4 postnatal weeks with WM maturation when controlling for birth weight, gestational age at birth and time between birth and scan.

Results: 91 VPIs were identified for our analyses after applying our inclusion criteria. After controlling for gestational age at birth and time between birth and scan, higher birth weight Z-scores were associated with greater WM tract maturation (as evidenced by greater fractional anisotropy and reduced diffusivity), most prominently in the corpus callosum and sagittal striatum. The postnatal changes in weight Z-score from birth to 4 weeks of life were also associated with increased maturity in these WM tracts, after controlling for gestational age at birth, birth weight Z-score, and time between birth and scan.

Conclusion: In VPIs, both higher birth weight and post-natal weight gain are independently associated with markers of brain WM maturation, particularly in corpus callosum, which can be captured on discharge MRI. Thus neuroimaging metrics of corpus callosum maturity on DTI can serve as potential biomarkers to examine early effects of nutritional interventions on VPIs’ brain development.