LAUSR

We are in agreement with the caveats put forward by Drs. Shahim and Diaz-Arrastia regarding our article.1 It is entirely possible that the MRI methods put forward in this paper are so sensitive that they detect changes that are of no functional consequence now. It is also possible that cognitive reserve in these individuals is sufficiently high that these changes have no consequence in the future. However, one could imagine that every life event that chips away at the brain's capacity for recovery and plasticity can ultimately affect the brain's response to a later-in-life insult, such as stroke or plaque formation. It is simply unknown whether this is a linear process or one in which a threshold needs to be surmounted, and hence, whether these are important or trivial changes. We would not advocate that these MRI approaches be used in a diagnostic manner. As noted in the comment, these are sophisticated and expensive approaches and are designed to study populations and inform directions for further research (and perhaps policy). One such direction is the use of blood biomarkers. We do have additional publications with data on this cohort in preparation but can note in passing that GFAP showed no changes at the sensitivity threshold of our techniques. In mild TBI or asymptomatic participants, metabolomic signatures may be more relevant than the classical markers such as GFAP or NFL, as we have previously noted.2 These would be more appropriate as accessible screening tools once we understand their relationship to the imaging results and perhaps more incisive cognitive testing.