Twenty carefully screened volunteers followed a sleep schedule that would be punishing in any workplace and especially risky in a spacecraft. Over 45 days, they lived through simulated missions that allowed only five hours in bed on weekdays and eight hours on weekends. By the end, their performance on a reaction-time test had dropped, even though their own fatigue ratings did not decline in the same clear way.
The study points to a problem that mission planners already worry about in low Earth orbit and will worry about even more on trips to the Moon or Mars. People can remain disciplined, motivated and outwardly functional while chronic sleep restriction quietly erodes alertness. In an operational setting, that means the first warning sign may come from a performance test rather than from a crew member saying they feel much worse.
A Scientific Reports study tracked 20 individuals across five simulated missions and found that psychomotor vigilance task performance declined from the start to the end of the mission. The paper also tested several bio-mathematical models that are supposed to forecast performance during sleep loss. Those models captured the average downward trend, but they were less reliable when the question shifted from group behavior to which specific person might be in trouble.
What the 45-day mission actually tested
The volunteers were described as astronaut-like because they had been rigorously selected for a demanding mission analogue. NASA’s Fatigue Countermeasures Laboratory lists the study among its HERA work on chronic sleep restriction, crew alertness and performance. The setting was meant to examine how a small crew functions when sleep is limited over many weeks, not to recreate every physical stress of orbital or lunar flight.
Researchers measured performance with the PVT, a reaction-time task widely used in sleep research because even brief lapses become visible when people are tired. They also collected Samn-Perelli fatigue ratings, which are self-reports of how tired a person feels. The contrast between those two measures gave the article its central tension: objective performance worsened over the mission while subjective fatigue changed much less clearly.
The paper compared weekdays after five-hour sleep opportunities with days that followed the longer weekend schedule. Crewmembers performed worse after the shorter nights, which fits decades of sleep science, but the mission design added something more realistic than a one-night deprivation experiment. It showed what happens when sleep debt accumulates inside an operational routine, with repeated work days, repeated testing and only partial recovery on weekends.
Why vigilance fell before self-report gave a clear warning
The clearest outcome was a progressive loss of alertness. The authors reported that performance declined from mission start to mission end and they also saw worse results on days after five-hour sleep periods than on days after eight-hour opportunities. On a simple test, that kind of drop means slower reactions and more lapses, the kind of small failures that can spread into larger mistakes when a crew is juggling procedures, checklists or monitoring tasks.
Fatigue ratings told a softer story. The crew rarely rated themselves at the most exhausted end of the scale, even though the performance data showed measurable deterioration. That gap is operationally important because many mission decisions still depend on a person’s own sense of readiness. When a crew member feels only moderately tired, a supervisor might assume the risk is manageable, while a reaction-time test may already be showing a sharper decline.
The paper put the practical lesson in blunt terms, writing that “sleep should be prioritized in lunar crews to minimize the potential for performance errors.” That line came from the study itself rather than from a press office summary and it fits the results closely. The issue was not a dramatic collapse. The issue was a steady slide in a high-performing group that still looked resilient on the surface.
Why the prediction models struggled with individual weak points
Sleep researchers often use mathematical models to estimate how much performance will drop after restricted sleep, circadian disruption or long wake periods. In this study, the models did a respectable job with averages across the mission. They reproduced the broad direction of change by day of mission and by time of day, which means they still have value for planning schedules and estimating general risk.
The weakness appeared at the individual level. Some crewmembers tolerated the schedule better than others and some showed concerning impairment early in the mission. The models were much less sensitive to those person-to-person differences. A planner using only the model output might see that the crew as a whole was drifting downward without knowing which individual needed closer monitoring or a change in workload.
That is one reason the result reaches beyond spaceflight. The paper noted that similar issues appear in other high-performing groups such as physicians, pilots and special operations units. Average predictions can still be useful, but they do not remove the need for direct measurement. For real missions, that supports a layered approach in which alertness testing, sleep scheduling and operational judgment are used together rather than treating a model as a final fitness-for-duty answer.
Why NASA treats sleep loss as a mission risk, not a comfort issue
NASA’s sleep risk overview says astronauts must maintain a high level of cognitive performance throughout a mission and notes that sleep deprivation and circadian disturbance are common in spaceflight. The agency also points to environmental factors such as noise, temperature, vibration and light as barriers to good sleep. The 45-day analogue study fits directly into that concern because it isolates one of the most basic drivers of error: insufficient time to sleep.
Another NASA page on behavioral health risk explains that prolonged isolation and confinement can affect sleep, morale and decision making, especially when communication delays become part of the mission. A crew on a Mars route will not have instant support from the ground and a sleep-related performance slump may appear during exactly the kind of delayed, autonomous operations that require sharp attention. Under those conditions, small lapses can carry more weight than they would in a short mission close to Earth.
The broader hazard frame also matters. NASA’s page on isolation and confinement says sleep loss, circadian desynchronization and work overload compound the strain of being cut off in a small habitat. The HERA-style result therefore belongs to a larger systems problem. Sleep is tied to scheduling, lighting, workload, privacy and habitat design, so protecting performance requires more than telling crews to rest when they can.
What the study can guide and what it still cannot prove
The strongest conclusion is narrow and useful. A rigorously selected group showed worsening objective performance during a 45-day mission analogue built around sleep debt and the usual prediction tools were better at crew averages than at person-specific vulnerability. That gives mission designers a credible reason to preserve sleep opportunity, monitor alertness directly and be cautious about assuming that elite selection alone will protect a crew from chronic fatigue.
The study also had limits that the authors acknowledged. It relied on the PVT as the main measure of impairment and poor PVT performance does not map perfectly onto every complex task in a real spacecraft. The mission analogue also focused on sleep restriction inside confinement rather than the full mix of microgravity, radiation, true distance from Earth and emergency demands that astronauts would face beyond low Earth orbit. Those limits keep the finding in the right frame.
Even with those caveats, the article offers a clear operational warning. A crew can look steady while reaction time grows less reliable and a scheduling model can look sensible while still missing the crewmember who is deteriorating fastest. For planners thinking about lunar bases, long HERA campaigns or eventual Mars flights, that argues for treating sleep opportunity and individual fatigue monitoring as mission hardware in all but name.
