You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster D22, Thursday, November 9, 6:15 – 7:30 pm, Harborview and Loch Raven Ballrooms

Can microstructural properties of cerebellar pathways improve prediction of reading skills in children?

Lisa Bruckert1, Katherine E. Travis1, Michal Ben-Shachar2, Heidi M. Feldman1;1Stanford University, 2Bar Ilan University

Introduction: Cerebellar pathways have been recently associated with reading abilities in typically developing children and adolescents (Travis et al., 2015). Children born preterm (PT) are at-risk for reading impairments and for cerebellar injury (Aarnoudse-Moens et al., 2009; Limperopoulos et al., 2010). Therefore, they represent an interesting group for studying the contribution of the cerebellum to reading. Here, we set to determine whether variation in microstructural properties of the cerebellar peduncles at age 6y improves the prediction of reading proficiency at age 8y, beyond socioeconomic status (SES) and behavioral measures, in PT and full term (FT) children. Methods: Thirty-eight children born FT (mean gestational age (MGA) = 39.6 weeks, 14 male) and 32 born PT (MGA = 29.8 weeks, 21 male) were assessed at age 6y with diffusion MRI (dMRI) and standardized tests of phonological awareness and reading, language, and non-verbal IQ. Fluency, comprehension and oral reading at age 8y were used as outcome measures. In dMRI, we applied a dual-spin echo diffusion-weighted sequence (3T, 30 diffusion weighted directions at b = 1000 s/mm2, 3 volumes at b = 0, voxel size 2 x 2 x 2 mm3). We used the Automated Fiber Quantification package (Yeatman et al., 2012) to segment whole brain deterministic tractograms and calculate the mean fractional anisotropy (FA) of the inferior (ICP), middle (MCP) and superior cerebellar peduncles (SCP) in each child. Correlations between mean FA and outcome measures were used to select cerebellar pathways included in subsequent analyses. Hierarchical multiple regression models assessed the contribution of SES, behavioral measures and mean FA at age 6y in predicting reading scores at age 8y. Results: Children born FT and PT did not differ significantly in outcome measures assessed at age 8y. Significant but weak associations were found between the mean FA of the left ICP and measures of fluency (r = 0.27; p = .039, N = 61) and oral reading (r = 0.29; p = .020, N = 66) at age 8y. A moderation analysis revealed that birth-group status significantly moderated the relationship between mean FA of the left ICP and the outcome measures. Multiple regression analyses were therefore carried out separately in the FT and PT groups. In children born FT, but not PT, mean FA of the left ICP predicted reading fluency (∆R2 = 0.10, F(1,30) = 6.83, p = .014) and oral reading (∆R2 = 0.13, F(1,30) = 11.5, p = .002) at age 8y, above and beyond SES, non-verbal IQ, language and phonological awareness. A parallel analysis of the left SCP did not reveal a significant contribution to prediction of reading outcome at age 8y. Conclusion: Microstructural properties of the left ICP (but not SCP) improved the prediction of reading skills at age 8y over known environmental and behavioral contributing factors. This was observed only in FT children, and not in PT children. These results suggest that children born premature may rely on alternative routes to achieve fluent reading.

Topic Area: Language Development

Back to Poster Schedule