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

Poster C33, Thursday, November 9, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Structural Brain Differences in Good and Poor Comprehenders Identified through a Regression-Based Quantitative Method

Kayleigh Ryherd1, Clint Johns2, Andy Jahn2, Julie Van Dyke2, Landi Nicole1,2;1University of Connecticut, 2Haskins Laboratories

Specific Reading Comprehension Disorder (S-RCD) is characterized by poor reading comprehension despite intact decoding ability (for review see Landi & Ryherd, 2017). To date, only one study has examined neurostructural differences related to S-RCD. Bailey et al. (2016) reported reduced gray matter volume (GMV) in right prefrontal cortex for S-RCD relative to typically developing (TD) readers, consistent with previous findings of impaired executive function in S-RCD readers. However, like most studies of S-RCD, Bailey and colleagues used a cutoff-based approach to determine group classification. Under this approach, readers classified as S-RCD must exhibit a discrepancy in their performance on standardized measures of word decoding and reading comprehension. Specifically, they achieve a standard reading comprehension score below some criterion, coupled with standard decoding performance that meets or exceeds an age-appropriate norm. Consequently, because this approach relies on dichotomizing continuous variables, individuals with quantitatively similar scores (i.e., scores close to the cutoff criterion) are treated as qualitatively different. Further, this practice is also associated with numerous statistical problems, including distorted effect sizes, reduced power, and increased Type 1 error (MacCallum et al., 2002). To address these weaknesses, some researchers have adopted a regression-based method to identify groups of readers differing on comprehension ability, but not on subcomponent measures (e.g., Tong et al., 2011, 2013). Under this approach, reading comprehension ability is predicted from theoretically justified component skills, e.g., nonverbal IQ, decoding ability, and vocabulary knowledge. Readers are classified as S-RCD if their comprehension scores are much lower than their predicted scores; TD if comprehension is commensurate with model predictions; and as unexpected good comprehenders (UGC) if their comprehension scores are much higher than predicted. Thus, this approach tailors group selection criteria to the individual by comparing reading comprehension skill to a predicted value based upon the individual’s known cognitive abilities, rather than relying on population norms. Our investigation is the first to capitalize on these advantages in an investigation of potential relations between comprehension skill and brain structure. In this study, we assessed cortical structural differences between three groups of comprehenders (TD, UGC, S-RCD), classified via the regression-based quantitative method. We predicted reading comprehension scores using age, decoding ability, nonverbal IQ and vocabulary. TD readers' comprehension scores fell within the 15% CI around the regression line. S-RCD readers had scores below the lower 65% CI of the regression line, while UGC readers' scores were above the upper 65% CI. From a sample of 172 participants (ages 13-24), this method identified 20 TD, 22 S-RCD, and 25 UGC. Exploratory whole-brain analysis of GMV revealed group differences, in which TD readers showed reduced GMV relative to UGCs in the left superior parietal lobule (SPL), a region which has been functionally implicated in improved comprehension ability following remedial instruction (Meyler et al., 2008). Our findings highlight the importance of studying UGCs (who, like S-RCD readers, have discrepant comprehension and decoding profiles) in addition to TD and S-RCD readers to gain a broader perspective on how brain structure may relate to reading comprehension skill profiles.

Topic Area: Language Disorders

Back to Poster Schedule