Angular Gyrus Recruitment During Inference Tracks Developmental Differences in Reasoning

• In Event: Poster Session 14
  In Poster Session: PS 14 Topic - Biological Processes: Neuroscience and Genetics

Abstract

Although the ability to infer new knowledge based on multiple memories improves into adolescence, little is known about the neurocognitive mechanisms contributing to this developmental trajectory. Prior work in adults indicates that the angular gyrus within the posterior parietal cortex plays a key role in both memory retrieval (Rugg & Vilberg, 2013) and reasoning (Seghier, 2013). Developmentally, posterior parietal activation tracks age-related improvements in memory through late childhood (Ghetti & Bunge, 2012), a period during which the angular gyrus approaches structural maturity (Seghier, 2013). Here, we examined whether angular gyrus function further tracks age-related improvements in memory-driven inference.

Children (7-12 years) and adults (N=74) completed an associative inference task (Figure 1A). They first learned an initial set of pairs (AB) to a high accuracy criterion. They then completed two additional learning phases (order counterbalanced), consisting of two study rounds and a scanned test each. In the overlapping pair phase, participants learned new pairs (BC) that shared one item in common with the initial pairs (B). In the non-overlapping learning phase, they learned pairs that contained only new items (XY). In a final scanning phase, participants completed a surprise inference test about the indirectly related items (AC) that shared a common associate (B). We predicted age-related improvements in performance would be greater for inference than memory, and that angular gyrus engagement during inference would track developmental differences in reasoning performance.

Mixed-effects models revealed all age groups performed similarly on learning non-overlapping associations (p=.069). In contrast, both adults and older children (10-12 years) performed better than younger children (7-9 years) on the inference and overlapping memory tests (ps<.001). We further found that parietal cortex (Figure 1B), and specifically angular gyrus (Figure 1C), was engaged more during inference (AC) than memory tests (BC/XY), with this difference in activation increasing with age even when controlling for reaction time. Angular gyrus engagement during inference also showed a differential relationship to performance between age groups, F(1,59)=6.18, p=0.016 (Figure 1D). In adults, angular gyrus activation facilitated inference performance, with greater activation relating to faster inference response times. In contrast, engagement at younger ages related to slower inference decisions.

Behavioral findings suggest overlapping experiences pose increased memory demands, which younger children resolve differently than adults. While adults and older children link prior experiences (AB) with new events (BC), younger children may not, which could limit their ability to infer relationships among indirectly related elements (AC) during learning or at the time of inference. Neural data indicate that developmental differences in memory-guided inference arise partially from functional maturation of angular gyrus. Work in adults (Morton et al., in press) indicates that posterior parietal cortex supports inference by efficiently navigating shortcuts through a cognitive map that links indirectly related events in memory. Here, angular gyrus activation in adults may lead to faster inference decisions via a similar process that may be unavailable at younger ages. More broadly, findings indicate that functional development of angular gyrus may play a key role in not only age-related increases in memory, but also reasoning.

Authors

• Christine A Coughlin, University of Texas at Austin
  Presenting Author

• Meg Schlichting, University of Toronto
  Non-Presenting Author

• Katherine R. Sherrill, University of Texas at Austin
  Non-Presenting Author

• Michelle M. Moreau, University of Texas at Austin
  Non-Presenting Author

• Alison Preston, University of Texas at Austin
  Non-Presenting Author