# Sixteen astronaut-aged volunteers worked four overnight space robotics sessions and low-dose caffeine improved vigilance while blue-enriched light shifted some EEG signs of sleepiness but still left later sleep under strain

> Sixteen astronaut-aged volunteers went through a demanding lab schedule that copied part of spaceflight life: a week of short sleep, sudden flips into overnight work and repeated robotics sessions during the biological night. In that setting, the clearest boost came from low-dose...

Canonical URL: https://www.argo.net/sixteen-astronaut-aged-volunteers-worked-four-overnight-space-robotics-sessions-and-low-dose-caffeine-improved-vigilance-while-blue-enriched-light-shifted-some-eeg-signs-of-sleepiness-but-still-left-l/
Byline: ARGO.net Editorial Team
Published: 2026-08-04T22:05:02+00:00
Categories: Explainer, Space

![A mission control room filled with monitoring screens](https://www.argo.net/wp-content/uploads/2026/08/space_mission_control_console.jpg)

Sixteen **astronaut-aged volunteers** went through a demanding lab schedule that copied part of spaceflight life: a week of short sleep, sudden flips into overnight work and repeated robotics sessions during the biological night. In that setting, the clearest boost came from **low-dose caffeine**. The stimulant improved reaction speed, cut long lapses in attention and helped people say they felt more alert.

The study, published in [**npj Microgravity**](https://www.nature.com/articles/s41526-023-00332-w) on December 19, 2023, also tested **blue-enriched white light** that was modeled on lighting developed for the **International Space Station**. Light alone helped on one information-processing task and nudged some brain-wave markers in a more alert direction, yet it did far less than caffeine on the main vigilance measures. The same paper also reported an important cost: caffeine made later sleep lighter and more broken.

## How the overnight spaceflight test worked

Researchers from **Brigham and Women's Hospital**, NASA-linked collaborators and other partners enrolled 17 healthy adults between ages 26 and 55, although one left early for a family emergency, leaving 16 complete data sets. Everyone stayed in a time-isolated inpatient unit for 13 days, followed controlled sleep schedules before entering the lab and gave up caffeine, alcohol, nicotine and other outside stimulants during the protocol.

Each participant first went through a baseline condition under standard white light and placebo pills. After that, the same people cycled through three countermeasure conditions: white light plus caffeine, blue-enriched light plus placebo and blue-enriched light plus caffeine. Because the design was within-subject, each person served as his or her own comparison point. The team also timed analysis to each volunteer's **dim light melatonin onset**, a marker of biological night, so the results would focus on the hardest part of the circadian schedule.

The work blocks were not simple button-press tests. Participants completed spaceflight-style robotics tasks and then moved into a cognitive battery that included the **Psychomotor Vigilance Test**, the digit symbol substitution task, alertness ratings and the **Karolinska Drowsiness Test**, which paired quiet wakefulness with EEG recording. A companion [space teleoperation paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC10730832/) from the same research program examined how these countermeasures affected spare visual attention during simulated robotic work.

## Why caffeine helped more than light alone

Baseline performance showed that the protocol worked as intended. After a week limited to six hours of sleep per night, participants were already impaired before the extra overnight strain was added. Their average reaction time on the vigilance task was 567 milliseconds and they averaged 12 lapses longer than 500 milliseconds, levels the authors compared with poorer performance than well-rested adults usually show.

Caffeine changed that pattern in a consistent way. Under white light plus caffeine, average vigilance speed rose from 2.7 to 3.2. Under blue-enriched light plus caffeine, it rose to 3.3. Lapses also dropped sharply, from about 12 at baseline to 5.1 with caffeine under standard light and 3.6 with caffeine under blue-enriched light. Subjective ratings moved in the same direction and sleepiness scores on the Karolinska scale fell from 6.0 at baseline to 4.5 and 4.3 in the two caffeine conditions.

Blue-enriched light by itself did not produce the same broad lift. It did not significantly improve vigilance speed, did not significantly reduce lapses and did not clearly improve self-rated alertness. It did help on the digit symbol substitution task, where all three experimental conditions beat baseline. That pattern suggests the lighting change may have supported some faster information processing even though it was too weak, by itself, to rescue the most fatigue-sensitive attention measures during this heavy schedule.

## What the EEG signals showed

Brain-wave data added another layer to the story. During the Karolinska Drowsiness Test, the team examined power in delta, theta, alpha and high-alpha bands, which are often used as objective clues about alertness and sleep pressure. Caffeine reduced delta power and theta power relative to baseline and both caffeine conditions also reduced alpha and high-alpha power. In plain terms, the EEG changes lined up with the behavioral result that people were staying more alert.

Light alone had a narrower signal. The 6300 K blue-enriched condition significantly reduced theta power and delta power moved in the same direction without reaching statistical significance. The authors treated that as evidence that the light had some alerting effect, though a moderate one. An earlier [blue light and caffeine study](https://pmc.ncbi.nlm.nih.gov/articles/PMC3838207/) also found that the two countermeasures can influence psychomotor function in different ways, which fits the split result seen here between vigilance and information processing.

Process details matter here. The light exposure was kept near 90 lux at the eye and was designed to remain workable for real task lighting rather than deliver a very strong blue stimulus. The paper argues that this practical choice probably limited the size of the light-only effect. Screen glow from the robotics workstation also narrowed the difference in melanopic light between the standard and blue-enriched conditions, which may have made the two lighting setups less distinct than they look on paper.

## Why the sleep tradeoff matters

Sleep after the work shift is where the clean success story stops. The percentage of wake during the sleep episode rose from 8.6 percent at baseline to 22.1 percent in the standard-light caffeine condition and 26.4 percent in the blue-light caffeine condition. In other words, the same countermeasure that kept people functioning better overnight also kept them awake more once they were supposed to recover.

The study also found less **slow wave sleep**, the deepest sleep stage, after every experimental condition, with the largest drops in the caffeine arms. Baseline slow wave sleep averaged 24.9 percent of the sleep episode. That fell to 14.5 percent with caffeine under standard light and 15.5 percent with caffeine plus blue-enriched light. Blue-enriched light alone also reduced slow wave sleep, though much less, to 21.4 percent.

Those numbers are important for operations planning because overnight alertness is only one part of mission safety. If a countermeasure keeps a crew member sharper during a difficult shift but cuts recovery sleep afterward, the gain may shrink over repeated days. The authors therefore suggested a more strategic use pattern, with caffeine ending around the midpoint of an overnight shift so some benefit remains for performance while less of the drug is still circulating near bedtime.

## What the results can and cannot tell NASA

Several parts of the experiment were unusually realistic. Participants were in the astronaut training age range, the robotics work was spaceflight-relevant and the lighting hardware was a prototype tied to the LED systems later installed on the ISS. A 2024 [NASA poster](https://ntrs.nasa.gov/api/citations/20240004245/downloads/2024%2003%2011%20%20AsMA%20Poster11402.pdf) on caffeine use aboard the station underlines why this question matters in practice: caffeine is already part of daily life in orbit.

Even so, the paper does not show how active astronauts would respond during a mission. The volunteers were healthy and carefully selected, but they were still ground-based participants in a simulation. The baseline condition always came first, which may have influenced later comparisons. The researchers also said the light level was probably too low to deliver the strongest alerting effect and they noted that performance under every countermeasure still remained poorer than what rested people usually achieve.

For NASA and for night-shift teams on Earth, the most careful reading is a limited one. Strategic caffeine looked like the more dependable short-term tool for maintaining alertness during biological night work. Blue-enriched light showed promise, especially in EEG and one processing task and might do more at higher melanopic levels that modern station lighting can provide. Neither tool solved the deeper problem created by chronic short sleep and circadian misalignment, which means schedule design still carries much of the safety burden. A [PubMed record](https://pubmed.ncbi.nlm.nih.gov/38114500/) for the paper and the full article both point to the same bottom line: countermeasures can recover part of the deficit, not all of it.