Substantial individual differences exist in how acute stress affects large-scale neurocognitive networks, including Salience (SN), Default Mode (DMN) and Central Executive Networks (CEN). These network-level changes upon acute stress may predict vulnerability to long-term stress effects, which can only be tested in prospective longitudinal studies. Using a longitudinal design, we investigated whether the magnitude of acute-stress induced functional connectivity changes (delta-FC) predicts the development of posttraumatic stress disorder (PTSD) symptoms in a relatively resilient group of young police recruits that are known to be at high risk for trauma-exposure. Using resting-state fMRI, we measured acute-stress induced delta-FC in 190 police recruits before (baseline) and after trauma exposure during repeated emergency aid services (16-month follow-up). Delta-FC was then linked to the changes in perceived stress levels (PSS) and post-traumatic stress symptoms (PCL and CAPS). Weakened connectivity between the SN and DMN core regions upon acute stress induction at baseline predicted longitudinal increases in perceived stress level but not of post-traumatic stress symptoms, whereas increased coupling between the overall SN and anterior cerebellum was observed in participants with higher clinician-rated PTSD symptoms, particularly intrusion levels. All effects remained significant when controlling for trauma exposure-levels and cortisol stress-reactivity. Except these neural effects, neither hormonal nor subjective measures were relevant. The reconfiguration of large-scale neural networks upon acute stress induction is relevant for assessing and detecting risk and resilience factors for PTSD. This study highlights the SN connectivity-changes as a potential marker for trauma-related symptom-development, which is sensitive even in a relatively resilient sample.

Posttraumatic stress disorder (PTSD) is a complex psychiatric condition that has generated much attention in the neuroimaging literature. A neurocircuitry model supporting fronto-limbic dysfunction as a major player in facilitating clinical symptoms of PTSD is well-characterized; however, recent literature suggests that network-based approaches may provide additional insight into neural dysfunction in PTSD. Our analysis uses resting-state neuroimaging scans of 1063 adults from the PGC-ENIGMA PTSD Consortium to investigate a network-based model of functional connectivity in PTSD. With a novel, resolution limit-free community detection approach, 16 communities corresponding to functionally meaningful networks were detected with high quality. After group-level community detection, participants were classified into three groups (PTSD, n=418, trauma-exposed controls without PTSD, n=434, and non-trauma exposed healthy controls, n=211). Individual network connectivity metrics were calculated, including whole-brain, default mode network, and central executive network participation coefficient and connectivity strength. Linear mixed effects models revealed group differences in the whole-brain, default mode, and central executive network participation coefficient and connectivity strength such that individuals with PTSD demonstrated overall greater values. We also described sex differences such that males demonstrate greater whole-brain participation coefficient vs. females and females demonstrate greater default mode network connectivity strength vs. males. Our results suggest that PTSD in adults is associated with reduced specialization and enhanced inter-module communication throughout the brain network, which may contribute to inefficient information processing and poor emotional regulation. This study presents a novel use of resolution limit-free community detection in a large PTSD sample, revealing robust differences in resting-state network topology.

No abstract