Abstract / Summary
Abstract Objective Late Central Nervous System (CNS) effects are currently under surveillance by NASA as a possible long-term health consequence of space radiation. Despite decades of follow-up, the small size of the astronaut cohort and long latency of these conditions have made conventional risk estimation difficult, leaving a gap in evidence to support deep-space missions to the Moon and Mars, where radiation exposures will exceed those of low Earth orbit. Methods We use data on NASA astronauts from the Astronaut and Spaceflight Participant Archive database linked to mortality information from the National Death Index to calculate standardized mortality ratios from late CNS effects from 1959 to 2022. We introduce a Bayesian surveillance system that determines how many deaths in the future are required to obtain various levels of probability for reductions or increases in mortality risk. Results Standardized mortality ratio (SMR) point estimates are below parity for both men and women. For males, SMR = 40 (95% CI = 5–145). For females SMR = 0 (95% CI = 0–4611) with only 0.08 expected. The surveillance system shows that, over 10 years, 17 total deaths for males and 3 for females would be required to achieve a 97.5% probability that SMR > 100. Conclusions The confidence intervals for both SMRs convey considerable uncertainty in the estimates. The Bayesian framework transforms this uncertainty into prospective surveillance thresholds, quantifying how many future deaths would signal excess or reduced risk and providing a transparent, updatable tool for monitoring rare outcomes in small cohorts.