Longitudinal brain-age predictions comprising long-duration spaceflight missions

Abstract

Our study investigates the effects of long-duration spaceflight on brain aging in spacefarers using structural MRI and machine learning models. Pre-, post-, and follow-up scans of ROS cosmonauts ESA astronauts, and matched Earth-bounding controls were analyzed. We found a considerable difference between the spacefareres and the control group, especially in the ESA cohorts (ß = 0.63). In the ROS cohorts, we observed a difference between the pre- and post-flight scans. A post-hoc analysis revealed that the pre-flight brain age delta was 0.842 years less than the immediate post-flight brain age delta after long-duration spaceflight. All three machine learning models showed good to excellent intraclass correlation coefficients (ICC) between the two consecutive MRI sessions. Our findings suggest that long-duration spaceflight may have an effect on human brain aging as observed from MRI.