Research: How Space Medicine Evidence Works

Why the evidence is thin

Most spaceflight studies follow only a handful of astronauts.

• Small numbers. A 2014 review found only 11 long-duration spaceflight studies of cognition, with 42 people in total [5]. A large exercise analysis had 46 long-duration astronauts [6]. The landmark NASA Twins Study had one astronaut and his identical twin on the ground [8].
• Mostly low Earth orbit (LEO). Nearly all long-duration data come from the International Space Station (ISS), inside Earth’s magnetic field. NASA notes that little is known for missions beyond LEO or in partial gravity [1][2].
• No randomized trials of spaceflight itself. Crews are their own controls: measurements before flight, during flight and after landing. Countermeasures are rarely compared head-to-head in flight.
• Many things change at once. Weightlessness, radiation, isolation, altered sleep and CO₂ all act together, which makes it hard to say what caused what.

Where the evidence comes from

See Earth analogs for what each analog is good and bad at.

How NASA rates risk

NASA’s Human Research Program (HRP) tracks a list of human-system risks grouped under five hazards of spaceflight: radiation, isolation and confinement, distance from Earth, gravity fields, and hostile or closed environments [1]. For each risk and each design reference mission, the Human System Risk Board assigns a likelihood × consequence (LxC) score on a 5×5 matrix and a disposition such as “Accepted” or “Requires Mitigation”. It also grades the level of evidence behind each risk [3]. The research needed to close each risk is broken into knowledge gaps in the Human Research Roadmap [2].

The Risk & Gap Atlas gives a plain-language page for each risk, with NASA’s current ratings and gaps.

Reading the certainty grades

SMF attaches a certainty grade to key claims, adapted from the GRADE approach used in clinical guidelines [4]. All grades on this site are AI-proposed and pending reviewer confirmation.

Evidence-type tags show where a claim comes from: F flight, A analog, M model/animal/cell, C computational, E expert/operational/standards.

Three tips:

1. A grade is about confidence, not importance. A risk can be serious and still have “low” certainty because it hasn’t been well studied.
2. A NASA rating can be “high” certainty even when the science is weak. The grade then describes how sure we are of what NASA’s rating says, not of the underlying biology.
3. Small studies swing. When a claim rests on fewer than 20 people, look for the individual range, not just the average. Many studies stress large differences between individuals [6].

A quick checklist for reading a space medicine study

Sections

• Risk & Gap Atlas: one page per NASA human-system risk.
• Landmark studies: the NASA Twins Study, Inspiration4 and SOMA, ISS cohorts, and bed rest.
• Earth analogs: what each analog is good and bad at.
• Datasets & resources: where to find the data.
• Open questions: the big unknowns, linked to NASA’s gaps.