Neural Correlates of Oral Word Reading, Silent Reading Comprehension, and Cognitive Subcomponents

Abstract

The ability to read is essential for cognitive development. To deepen our understanding of reading acquisition, we explored the neuroanatomical correlates (cortical thickness (CT)) of word reading fluency and sentence comprehension efficiency in Chinese with a group of typically developing children (N = 21; 12 females and 9 males; age range 10.7-12.3 years). Then, we investigated the relationship between the CT of reading-defined regions and the cognitive subcomponents of reading to determine whether our study lends support to the multi-component model. The results demonstrated that children's performance on oral word reading was positively correlated with CT in the left superior temporal gyrus (LSTG), inferior temporal gyrus (LITG), supramarginal gyrus (LSMG) and right superior temporal gyrus (RSTG). Moreover, CT in the LSTG, LSMG and LITG uniquely predicted children's phonetic representation, phonological awareness, and orthography-phonology mapping skills, respectively. By contrast, children's performance on sentence reading comprehension was positively correlated with CT in the left parahippocampus (LPHP) and right calcarine fissure (RV1). As for the subcomponents of reading, CT in the LPHP was exclusively correlated with morphological awareness, whereas CT in the RV1 was correlated with orthography-semantic mapping. Taken together, these findings indicate that the reading network of typically developing children consists of multiple subdivisions, thus providing neuroanatomical evidence in support of the multi-componential view of reading.

Keywords: individual differences; multi-component model; neuroanatomy; reading comprehension; reading fluency; word reading.

Figure 1

A. Results of the whole-brain regression analysis between word reading efficiency and cortical thickness, controlling for age, gender and performance IQ (n = 21). Clusters that survived the random field theory (RFT) correction are presented on a standard inflated surface template. Left panel: clusters with an RFT-corrected p < 0.05 are projected on a standard surface template. Right panel: scatter plots present the brain-behavior correlation in each significant cluster (x-axis: cortical thickness in mm, y-axis: reading scores adjusted for age, gender and performance IQ). B. Results of sentence reading comprehension are presented in the same way.

Figure 2

Scatter plots display the correlation between the significant predictor and the reading subcomponents (n = 21). The subcomponent scores were adjusted for age, gender and performance IQ. A. Phonological representation is predicted by cortical thickness in the left superior temporal gyrus. B. Phonological awareness is predicted by cortical thickness in the left supramarginal gyrus. C. Orthography-phonology mapping is predicted by cortical thickness in the left inferior temporal gyrus. D. Morphological awareness is predicted by cortical thickness in the left parahippocampus. E. Orthography-semantic mapping is predicted by cortical thickness in the right calcarine fissure.