As part of my doctoral dissertation for my PhD in Epidemiology from the University of Texas Health Science Center at Houston, I completed a three-phase project which asked whether the mortality rates for astronauts increase with the amount of ionizing radiation they have absorbed in space. Though a simple question to ask, answering it presented a challenge. In this collection of four posts I discuss why it was so difficult, how I got around those difficulties, and finally – the results!

Thanks to a prior study of astronaut mortality completed by myself and Steven Day, I started with a biographical database of all US astronauts selected between 1959 and 2009. I originally constructed this database by mining sources such as the Astronaut Fact Book and astronaut biographies on the Johnson Space Center website. For the purposes of my dissertation project, I first needed to update the database to include the class of astronauts selected in 2010, and to record any new deaths that had occurred since 2009. This was a relatively small part of the story, however; in order to determine if radiation exposure was leading to increased mortality among astronauts, I would need a reasonable measure of radiation exposure.

Ideally, dose measurements would come from individual dosimeter badges from each astronaut after each spaceflight. This would allow us to observe the mortality risk over time as a function of radiation dose, time since exposure, and related factors. Unfortunately those measurements, if they even existed, would be difficult to obtain, and would almost certainly have to come from the National Aeronautics and Space Administration (NASA). Knowing from experience how difficult it is to gain the cooperation of NASA researchers for a project like this, and owing to limitations on the time to complete my dissertation, I decided to use publicly available information to estimate radiation exposure on each mission. For that I would need a tool for estimating doses received on any given mission.

A review of the literature revealed a few papers with some rough estimates of radiation doses for mission profiles, but nothing specific enough to estimate the doses on a wide variety of missions. Instead I would have to create a new tool for dose estimation. With such a tool I could estimate doses for all manned missions, use those estimates for individual mission doses, and then model the relationship between the doses and mortality.

Thus Phase 1 of my project was to find the data I would need in order to estimate doses on manned missions. Phase 2 would use those measurements to create a dose estimation tool. Phase 3 would use the tool to estimate doses for all spaceflights, and then examine the relationship between cumulative radiation dose and mortality. Future posts (Parts 2-4) will summarize each of these phases and their results in turn. Of course, for a more complete telling of this story, one may read my dissertation at the University of Texas!