PURPOSE: To determine whether diffusion tensor imaging (DTI) and fractional anisotropy (FA) values in specific white matter (WM) tracts predict scores on a standardized assessment of motor skills at term and 12 wk, in a cohort of preterm infants

We hypothesized the DTI FA values in WM tracts would be lower in those infants with below average (at-risk) vs. average motor skills scores.

BACKGROUND: Infants with WM brain injury typically receive early intervention services well after maladaptive motor movements become fixed. Therapy services could lessen or prevent permanent deficits if infants could be identified earlier. The combination of motor testing and neuroimaging is a promising indicator of future impairment. DTI measures the directionality of water diffusion in the brain. FA gives information on microstructural integrity and axonal organization and is a sensitive marker of injury in WM.

DESIGN: A prospective study of preterm infants born between 24 and 34 wk of gestation (GA; N = 26). Exclusion criteria were infants born below 24 wk or infants with major congenital abnormality.

METHOD: Developmental testing was completed at term and 12 wk corrected at (CA) using the Test of Infant Motor Performance (TIMP). Scores were grouped into average (low risk) or below average (high risk) performance at 12 wk CA (TIMP manual). Neuroimaging data collection was performed at the Center for Biomedical Imaging at a mean age of 42.2 wk (standard deviation = 1.5) CA.

ANALYSIS: Voxelwise statistical analysis of the FA data was carried out using Tract-Based Spatial Statistics (TBSS) corrected for multiple comparisons. We removed volumes that had artifacts due to motion, and TBSS projected all participants’ FA data onto the Johns Hopkins University (JHU) neonatal FA atlas before applying voxelwise cross-subject statistics at the p < .05 significance level. Using the JHU neonatal atlas, we quantified mean FA in individual WM structures and created a linear model using DTI FA in individual WM structures, GA at scan, and GA at birth to predict scores on these individual TIMP items.

RESULTS: Significant differences were found between 12-wk TIMP high- vs. low-risk groups’ FA values in multiple regions including inferior longitudinal fasciculus (ILF), optical radiation (OR), inferior fronto-occipital fasciculus (IFOF), the corpus callosum (CC), the posterior limb of the internal capsule, and corticospinal tract (CST) (p < .05). Using a linear model including GA at birth and GA at scan, we found that FA of the left anterior limb of the internal capsule (ALIC) predicted mean change between term and 12-wk TIMP scores on head lift in prone (p = .045) as did the superior fronto-occipital fasciculus (p = .044). FA of the whole (p = .008) and left (p = .002) ALIC predicted mean change between term and 12-wk TIMP scores on left head turn in prone to sound.

DISCUSSION: Our results indicate that low FA values in specific WM tracts in the CC, CST, IFOF, ILF, and OR can predict overall motor skills at 3 mo. Specifically, several WM tracts identified through TBSS indicate that connectivity of the frontal cortex with the occipital visual cortex, the temporal lobe, and thalamus are important to early motor skills.

IMPACT STATEMENT: Understanding the relationship between brain development, early motor movements, and early language and visual development may lead to earlier identification of disorders and result in timely provision of early intervention services. We speculate sound may be used to motivate volitional movement associated with better head control. Therapy that targets visual– and auditory–motor intervention could improve WM tract injury due to preterm birth.