Schedule of Events | Symposia

Graduate Student Award Winner

Acute Stress Increases Functional Brain Network Integration

Poster Session D - Monday, March 31, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Chelsea C. Ajunwa¹ (ajunwa.c@northeastern.edu), Philip A. Kragel², Lawrence L. Wald^3,4, Marta Bianciardi^3,4, Tor D. Wager⁵, Ji-Kyung Choi⁶, Jiahe Zhang^1,4, Ajay B. Satpute^1,4, Karen S. Quigley¹, Lisa Feldman Barrett^1,4, Jordan Theriault^1,3; ¹Northeastern University, ²Emory University, ³Martinos Center for Biomedical Imaging, Massachusetts General Hospital, ⁴Massachusetts General Hospital and Harvard Medical School, ⁵Dartmouth College, ⁶University of California, San Francisco

Brain modularity refers to the brain’s functional organization into densely connected subnetworks. High modularity (i.e., low network integration) is thought to facilitate efficient and adaptable localized information processing; yet, within individuals, the brain can temporarily shift towards a more integrated (i.e., less modular) state, both during cognitive load and after acute social stress. We used 7 Tesla (1.1 mm isotropic) whole-brain imaging to examine brain modularity during (as opposed to after) acute social stress, and to quantify within-individual reliability of brain modularity (across multiple resting state scans). Participants (N = 92) were scanned during three consecutive resting state runs and a social stress task. For each participant and run, we constructed Pearson correlation connectivity matrices (using the 264 Power Atlas; correcting for multiple comparisons) and recorded the average modularity index (Q) across 100 iterations of the Louvain community detection algorithm. During acute social stress, brain modularity decreased from levels at rest (t = 6.377, p = 7.35 x 10-9). Modularity did not significantly differ across the three resting state runs (F = 0.969, p = 0.382). Within individuals, modularity was moderately reliable across resting state scans (interclass correlation coefficient of 0.486). At the same time, inter-individual differences in resting modularity were moderately large (rest SD = 0.147, stress SD = 0.173). In sum, acute social stress elicited reliable changes in modularity, and inter-individual differences were stable across scans. Additional planned analyses include examining cortical/subcortical differences in stability of network modularity, and correlating baseline and stress-induced modularity changes with behavioral performance.

Topic Area: EMOTION & SOCIAL: Other