Back

MRI signature of brain age underlying post-traumatic stress disorder in World Trade Center responders

Invernizzi, A.; La Rosa, F.; Sather, A.; Rechtman, E.; Jalees, M.; Nabeel, I.; Pellecchia, A.; Santiago-Michels, S.; Bromet, E.; Lucchini, R. G.; Luft, B. J.; Clouston, S.; Beck, E. S.; Tang, C.; Horton, M.

2024-10-19 occupational and environmental health
10.1101/2024.10.18.24315761 medRxiv
Show abstract

The men and women involved in rescue and recovery operations at the 9/11 World Trade Center (WTC) site have a greater prevalence (23%) of persistent, clinically significant post- traumatic stress disorder (PTSD). Recent structural and functional magnetic resonance imaging (MRI) studies demonstrate significant neural differences between WTC responders with and without PTSD. Here, we used brain age, a novel MRI-based data-driven biomarker optimized to detect accelerated structural aging, and examined the impact of PTSD on this process. Using BrainAgeNeXt, a novel convolutional neural network trained and validated on 11,574 magnetic resonance imaging (MRI) T1- weighted scans, we predicted brain age in WTC responders with PTSD (WTC-PTSD, n = 47) and age/sex matched responders without PTSD (non-PTSD, n = 52). Predicted Age Difference (PAD) was then calculated for each WTC responder by subtracting chronological age from brain age. A positive PAD indicates that the responders brain is aging faster than expected for their chronological age. We found that PAD is significantly greater with WTC-PTSD compared to non-PTSD responders (p < 0.001). Further, we found that WTC exposure duration (months working on site) moderates the association between PTSD and PAD (p=0.0050). Our results suggested that brain age is a valid biomarker to compare aging trajectories in responders with and without PTSD. In particular, PTSD may be a substantial risk factor for accelerated neurodegeneration in this vulnerable and aging population.

Published in Translational Psychiatry (predicted rank #1) · training set

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.