Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Developmental changes in neural states during speech planning for simple phrases in school-age children and adults
Poster Session A, Friday, September 12, 11:00 am - 12:30 pm, Field House
Marco Chia-Ho Lai1, Ellie Abrams1, Sherine Bou Dargham2, Jacqui Fallon3, Ebony Goldman4, Miriam Hauptman5, Alicia Parrish1, Sarah F. Phillips1,6, Alejandra Reinoso7, Liina Pylkkänen1,8; 1New York University, 2King’s College London, 3University of Colorado Boulder, 4City University of New York Graduate Center, 5Johns Hopkins University, 6Georgetown University Medical Center, 7Northwestern University, 8New York University Abu Dhabi
While previous behavioral studies indicate similar speech planning processes for language production in children and adults, neuroimaging studies have identified age-related differences in the neural mechanisms underlying speech planning during single-word production. However, less is known about how children and adults plan longer utterances, such as simple phrases. To address this, we used a picture naming task with magnetoencephalography (MEG) to investigate the speech planning of simple phrases in school-age children and adults. This study aimed to examine: (1) whether there are age-related changes in the neural mechanisms underlying speech planning during phrase production, and (2) when children begin to exhibit adult-like patterns of speech planning for phrases. Eighteen school-age children (9 females; mean age 9.7 ± 2.5 years) and 20 adults (13 females; mean age 26.7 ± 11.3 years), all native English speakers, participated in this study. During a picture naming task, participants described a picture using either an adjective-noun phrase (e.g., “green house”) for phrase naming or a noun (e.g., “house”) for word naming, while their brain activity was simultaneously recorded using MEG. For data analysis, we applied a hidden semi-Markov model (HsMM) to identify sub-states during speech planning by detecting state transitions through fitting a half-sine impulse model. Source activity per state was then estimated for subsequent comparisons. Behaviorally, both groups had lower accuracy and slower response times (ps < .001) during phrase naming compared to word naming. Across conditions, children showed lower accuracy and slower response times than adults (ps < .001). Sensor-level MEG results using the HsMM approach indicated that, during speech planning, the optimal number of events was three (yielding four states) for adults and four (yielding five states) for children. Source-level spatial clustering permutation tests revealed that, when contrasting phrase and word naming, adults showed greater activation in the right temporoparietal and posterior medial regions during state 2 (p = .008; 310 vertices; approximate 170-410 ms), followed by increased activation in the left insula and anterior temporal lobe (ATL) during state 3 (p = .035; 373 vertices; approximate 360-660 ms). Under the same contrast, children exhibited greater activation in the left parietal regions during the first two states, with a spatial shift from parietal to temporal areas between state 1 (p = .046; 181 vertices; approximate 0-240 ms) and state 2 (p = .031; 202 vertices; approximate 200-450 ms). Correlational analyses further showed that, in state 5 (approximate 750-1080 ms), an adult-like activation pattern in the left ATL during phrase naming gradually emerged with increasing age in children (Spearman’s rho = 0.62, p = .006). In sum, our study demonstrated distinct speech planning processes between children and adults during picture naming. Specifically, adults recruited bilateral regions related to feature binding and compositional process, with distinct temporal profiles, when planning simple phrases. In contrast, children engaged more attention- or lexico-semantic related left temporoparietal regions during the early stages of phrase naming. Furthermore, age correlations in children indicated a prolonged development of the left ATL to specialize in semantic composition for phrase production.
Topic Areas: Language Development/Acquisition, Language Production