White Matter Injury and General Movements in High-Risk Preterm Infants

Abstract

Background and purpose: Very preterm infants (birth weight, <1500 g) are at increased risk of cognitive and motor impairment, including cerebral palsy. These adverse neurodevelopmental outcomes are associated with white matter abnormalities on MR imaging at term-equivalent age. Cerebral palsy has been predicted by analysis of spontaneous movements in the infant termed "General Movement Assessment." The goal of this study was to determine the utility of General Movement Assessment in predicting adverse cognitive, language, and motor outcomes in very preterm infants and to identify brain imaging markers associated with both adverse outcomes and aberrant general movements.

Materials and methods: In this prospective study of 47 preterm infants of 24-30 weeks' gestation, brain MR imaging was performed at term-equivalent age. Infants underwent T1- and T2-weighted imaging for volumetric analysis and DTI. General movements were assessed at 10-15 weeks' postterm age, and neurodevelopmental outcomes were evaluated at 2 years by using the Bayley Scales of Infant and Toddler Development III.

Results: Nine infants had aberrant general movements and were more likely to have adverse neurodevelopmental outcomes, compared with infants with normal movements. In infants with aberrant movements, Tract-Based Spatial Statistics analysis identified significantly lower fractional anisotropy in widespread white matter tracts, including the corpus callosum, inferior longitudinal and fronto-occipital fasciculi, internal capsule, and optic radiation. The subset of infants having both aberrant movements and abnormal neurodevelopmental outcomes in cognitive, language, and motor skills had significantly lower fractional anisotropy in specific brain regions.

Conclusions: Aberrant general movements at 10-15 weeks' postterm are associated with adverse neurodevelopmental outcomes and specific white matter microstructure abnormalities for cognitive, language, and motor delays.

© 2017 by American Journal of Neuroradiology.

Figures

Fig 1.

Differences in white matter at term-equivalent age between infants with normal and aberrant fidgety movements at 10–15 weeks. Mean FA skeleton, in green, is overlaid on the mean FA map. Superimposed are pseudocolored voxels having significantly greater anisotropy in infants with normal than in infants with aberrant fidgety movements. Color bar shows the range of P values represented by the pseudocolors. Significantly higher regions of FA can be observed in the splenium (ccs) and genu (ccg) of the corpus callosum, inferior (ilf) and superior longitudinal fasciculus (slf), fronto-occipital fasciculus (fof), anterior (alic) and posterior (plic) internal capsule, corona radiata (cr), cerebellar peduncles (cp), and fornix/stria terminalis (fx/st). Z represents the MR imaging axial section coordinates (zero is the center of the anterior commissure).

Fig 2.

Differences in white matter between infants with “doubly normal” (normal fidgety movements and normal BSID-III scores) and “doubly abnormal” (aberrant fidgety movements and abnormal BSID-III scores) outcomes. The mean fractional anisotropy skeleton, in green, is overlaid on the mean fractional anisotropy map. Superimposed are pseudocolored voxels having significantly greater fractional anisotropy in “doubly normal” outcome infants than in infants with “doubly abnormal” assessments. The color bar shows the range of P values represented by the pseudocolors. Subscores on the BSID-III are divided into Cognitive (A), Language (B), and Motor (C) fractional anisotropy maps. Fiber tract labels are the same as in Fig 1.

Fig 3.

Differences in white matter in infants with aberrant fidgety movements between subpopulations with normal and adverse neurodevelopmental outcomes on the BSID-III. The mean fractional anisotropy skeleton, in green, is overlaid on the mean fractional anisotropy map. Superimposed are pseudocolored voxels having significantly greater fractional anisotropy in infants with normal than in those with adverse neurodevelopmental outcomes on Cognitive (in red-yellow) and Motor (in blue-light blue) BSID-III subscales. The color bar shows the range of P values represented by the pseudocolors. Fiber tract labels are the same as in Fig 1. Icp and mcp indicate inferior and middle cerebellar peduncles; ml, medial lemniscus.

Fig 4.

Mean fractional anisotropy in infants with normal and adverse neurologic outcomes. Boxplots represent the distribution of individual subject values in groups with normal fidgety movements and normal BSID-III scores, aberrant fidgety movements and normal BSID-III scores, and aberrant fidgety movements and adverse neurodevelopmental outcome as measured by BSID-III at 24 months of age. The cognitive subscale was measured in the splenium of the corpus callosum. The motor subscale was measured in the cerebellar peduncle. The asterisk indicates P < .05 on ANOVA, Tukey post hoc comparisons.