Can Modafinil or Galantamine Restore Simulator Flying After 27 Hours Awake? A 12-Person Crossover Study

Short answer: Not convincingly in this experiment. Twelve medically screened men completed three separate, randomized, double-blind sessions involving a single 100-mg oral modafinil dose, a single 10-mg oral galantamine dose, or placebo after prolonged wakefulness. The investigators observed some differences in simulated altitude control across visual environments and a selected galantamine-versus-placebo speed comparison. Yet no consistent drug advantage appeared across the three assessed flight-control parameters, and the data did not demonstrate a reliable restoration of rested flight performance. This was a small, fixed-base simulator experiment, not an actual flight, a chemical blood-drug study, or an individually validated test of aviation fitness.

The primary report is Lewkowicz R, Przewodzka A, and Łaszczyńska J, “Flight Performance Following Administration of an Anti-Fatigue Stimulant During 27 Hours of Continuous Wakefulness,” Polish Journal of Aviation Medicine, Bioengineering and Psychology, volume 28, issue 2, pages 5–18, with DOI 10.13174/pjambp.23.05.2025.01. The printed issue identifies 2022, while the DOI string includes 2025; those are distinct bibliographic elements and are not silently harmonized here. The practical question is what its directly measured simulator data can—and cannot—tell a forensic or operational reviewer.

What is the scientifically defensible bottom line?

Sleep loss affects multiple domains of performance, and a medication that promotes wakefulness need not normalize control of an aircraft. In this particular study, the overall pattern was limited rather than broadly restorative. The researchers tested altitude, speed, and heading maintenance in a simple flight scenario. They found no significant drug differences in altitude or heading within any given flight environment. One speed analysis yielded a difference involving galantamine, but most comparisons did not show a significant treatment effect. No convincing evidence established that a single modafinil or galantamine dose put participants back at a rested, pre-deprivation level across those outcomes.

That conclusion is intentionally narrower than either “the medications work for tired pilots” or “the medications can never work.” The tested doses, a particular sample, one simulator, and a specific prolonged-wakefulness schedule are the relevant frame. Several published statistical comparisons were null. A null result in twelve volunteers is not proof that an effect is impossible; it is a failure to demonstrate one under the tested conditions. Conversely, a within-condition positive contrast does not certify safe operation of a real aircraft or a specific driver’s or pilot’s fitness.

Who participated, and why does the true sample size matter?

The investigators enrolled twelve independent healthy male volunteers, ages approximately 20–25, with a reported mean age of 24 years. They were students at an academy of physical education in Warsaw and had medical clearance for the protocol. The paper does not describe them as a representative sample of active airline or combat pilots. Screening included clinical and physiological assessments. Medical exclusions included conditions that could increase the risks associated with the study medications. Every participant underwent three drug-condition sessions; this did not produce thirty-six independent men.

Each participant’s repeated flights, flight segments, telemetry points, and drug visits provide repeated observations on the same individual. Treating twelve people multiplied by three sessions as thirty-six statistically independent subjects would understate correlation and inflate evidentiary weight. The authors used repeated-measures nonparametric statistics for their treatment contrasts, which is consistent with the within-person design. This distinction is central when comparing the study to forensic databases with many genuinely independent drivers or blood specimens.

Which drugs were administered, at what doses and by what route?

At the appropriate test session participants received 100 mg modafinil orally, 10 mg galantamine orally, or an oral placebo. Modafinil is a wakefulness-promoting agent. Galantamine, a reversible acetylcholinesterase inhibitor, was included as a pharmacologically different candidate that could influence attention and task control. Calling these agents chemically equivalent or treating galantamine as an amphetamine would be incorrect. No methamphetamine or dextroamphetamine treatment arm was part of this study.

The authors used commercial oral preparations and a single dose per session. These are administered milligram quantities, not measured circulating concentrations. The trial did not directly assay modafinil, galantamine, their metabolites, or any stimulant parent drug in postdose whole blood, plasma, serum, urine, oral fluid, hair, or breath. It therefore provides no dose-to-concentration equation, threshold concentration for operational safety, or validated forensic pharmacokinetic curve. A known tablet dose and a confirmed specimen result are different categories of evidence.

How did the three-session crossover work?

The investigators described a randomized, double-blind, placebo-controlled crossover. Each person attended three nonconsecutive experimental sessions separated by about seven days, with the sequence of the two medications and placebo randomized. A session began in the evening and extended through the following afternoon. The paper described the medication administration at approximately 10:30 a.m., after a long period without sleep; the afternoon simulator test followed later, so the exact hours awake at medication and at the final simulator cannot be collapsed into a single universal exposure time.

The research protocol also included physiological monitoring, other psychomotor and flicker-detection tasks, and controlled heat exposure with light cycling activity. The publication analyzed the flight-simulator measurements rather than reporting a complete battery of all those additional outcomes. The latter tasks cannot be treated as hidden positive evidence for the drug unless their numerical results are actually reported. Separate sessions and familiarization flights helped address practice effects, but the possibility of residual learning across repeated simulated flights remains a methodological limitation.

What did the flight simulator directly measure?

Participants used a fixed-base Hyperion simulator and were instructed to maintain a nominal 200 meters above mean sea level, a nominal 400 kilometers per hour, and a specified heading of approximately 280 degrees. A typical simulated route lasted about 480 seconds. The route passed over land and water. Investigators extracted segments classified as mixed land-and-sea, land-only, and sea-only conditions and computed flight parameters from the simulation software.

These are virtual control variables: altitude in meters, speed in kilometers per hour, and heading in degrees. They were not chemical measurements, symptom questionnaires, or real crash outcomes. A stable heading or altitude within a simplified simulator is relevant to one component of flight control, but piloting also involves situational awareness, air-traffic communications, unexpected failures, navigation, divided attention, and judgment. The original protocol did not test those operational domains under actual airborne conditions.

Original evidence table: independent people, dosing and measurable endpoints

The measured outcomes were deviations or recorded values from those targets, categorized across visual flight environments. A target value should never be mislabeled as the group’s actual mean. Similarly, the brief description of a three-arm crossover does not imply three different cohorts; every volunteer contributed data under the separate drug conditions. These elementary distinctions are necessary before deciding whether a numerical change has any external meaning.

What did the statistical tests actually compare?

For the post-treatment flight test, the researchers compared medication conditions using the Friedman test, which handles repeated observations when normality assumptions are questionable. When indicated, they used Wilcoxon signed-rank contrasts with Bonferroni adjustment. They also examined flight environment—mixed land and sea, land, and sea—because visual reference availability differed. The reported Kendall’s W values describe within-sample association or effect size for a particular comparison; W does not directly express the percentage of pilots made safe by a drug.

Two kinds of comparisons must stay separate. A drug comparison within a terrain condition asks whether people under modafinil, galantamine or placebo controlled the variable differently in that scenario. A terrain comparison under one drug asks whether the same participant’s performance varied across visual environments while on that treatment. A significant terrain effect under galantamine is not, by itself, proof of galantamine superiority over placebo. Several summaries of this paper risk confusing these two questions.

Original evidence table: positive and null simulator results

These are heterogeneous comparisons. In particular, a significant change between two environmental scenes does not demonstrate a therapeutically useful drug-versus-placebo benefit. Conversely, the nonsignificant modafinil environment comparison at p = 0.053 remains nonsignificant under the specified p < 0.05 convention; calling it statistically significant because it is numerically close would change the authors’ results. Reported p-values also do not convey individual response distributions or clinical relevance.

Did modafinil improve altitude maintenance?

Not as a consistent drug effect shown within the separately analyzed flight scenes. The original Results specifically reported no statistically significant altitude differences between the three treatments for any single flight condition. The modafinil condition showed a trend toward different altitude distributions among the scenes, but its overall environment test had p = 0.053, above the conventional threshold. This does not establish a modafinil-specific benefit and should not be promoted as a positive drug finding.

The plotted altitude distributions provide descriptive context: participants tended to fly at different altitudes over feature-rich land and less-referenced water. Although an unadjusted visual impression may favor one condition on a particular panel, an inference about drug efficacy must follow the reported appropriate treatment test and its uncertainty. No detailed individual altitude values, blood exposure values, or repeatable aviation performance threshold was established that could assign a safe or unsafe status to an individual participant.

What did the investigators find with galantamine and visual terrain?

For galantamine sessions, altitude varied significantly across flight environments, with an overall reported χ²(2) = 10.9 and p = 0.004. The source described meaningful contrasts when sea and land environments were compared. Water scenes lacked some external visual references that would ordinarily help someone judge height. The experiment therefore contains an interaction between environmental difficulty and the person’s flight-control task, not a one-number drug effect that would transfer to every route or weather condition.

Galantamine was also associated with a lower simulated speed than placebo in a selected comparison, according to the source Results. Slowing toward a target could in some circumstances look like a cautious choice, but the report did not show that each subject consciously chose a safer speed or improved overall hazard management. Slower is not a universal synonym for better: flight may require maintaining a designated speed, not being slower than another experimental arm. The explicit instructed target and simulator context must anchor interpretation.

Was speed consistently improved by either medication?

No broad speed benefit was established. The one significant overall treatment comparison was reported for the mixed land-and-sea condition, with χ²(2) = 8.63 and p = 0.013; a post hoc galantamine-placebo contrast was reported as p < 0.001. Other condition and treatment analyses were not reported as demonstrating consistent statistically significant drug advantages. The paper described galantamine-associated slower speeds, but it did not establish an overall flight-safety utility function indicating how much slower is optimal.

The important endpoint is the instrument-recorded speed relative to the instruction, not a speculation that slowing automatically prevents accidents. Because the original figure material contains a labeling ambiguity that is handled in the separate editorial review, this analysis relies on the numerically reported inferential text rather than extracting precise speed means from a mismatched plotting scale. No unsourced kilometer-per-hour difference, stopping-distance calculation, or individual decision-time inference is supplied.

Did heading differ between the drugs?

The original report found no statistically significant heading differences among medications within any flight environment. A comparison of heading across environmental conditions during galantamine treatment produced an omnibus χ²(2) = 11.5, p = 0.003, but the authors stated that subsequent corrected pairwise comparisons did not identify significant differences. Reporting only the omnibus p-value as proof of a specific sea-versus-land improvement would ignore that subsequent null result.

Heading is only one control axis. It is sensitive to route geometry, simulator input devices, visibility, training, task simplicity and fatigue. The original plot presents broad within-condition variation. It does not establish that a person would safely manage an engine malfunction, approach, landing, instrument failure or competing communications during actual flight. The distinction between a simulator metric and a legally consequential operational conclusion remains essential.

What does “return to baseline” mean in this protocol?

The investigators compared earlier simulator performance with the later test after the medication period. They discussed their findings as failing to demonstrate full restoration of baseline accuracy. At the same time, the numerical Results reported no statistically significant differences in the selected earlier-versus-later comparisons across several analyzed parameters. A nonsignificant comparison cannot prove equality or restoration, just as it cannot alone prove persistent impairment. Formal equivalence requires a justified margin, adequate precision, and a design suited to that claim.

The earlier control time was itself during an evening period following a day of waking activity, rather than a scientifically established optimal rested reference for all subjects. The authors also acknowledged possible practice effects from repeated simulation. These design factors weaken simplistic statements about complete pharmacologic rescue or definitive failure. A careful formulation is that this small experiment did not establish reliable restoration of rested aviation performance following a single dose under its protocol.

Why do the wakefulness schedule and intervening tasks matter?

Participants completed a regular day before remaining awake overnight. The article described a median wake-up around 7:00 a.m. on the preceding day and medication the next morning. Thus, “27 hours awake” refers approximately to the exposure at dosing, not the number of awake hours at every later flight measurement. The afternoon assessment occurred after additional wakefulness. Timing differences influence the interpretation of expected drug action, fatigue, circadian phase and the absence of a measured blood concentration-time profile.

Between assessments, participants also underwent controlled thermal exposure near 30°C with cycling intervals and rest periods. That added exertion could affect fatigue and performance independently of the assigned pill. Their caffeine intake was restricted during testing, unlike many routine military contexts. The study therefore cannot isolate the effect of the drug from every aspect of prolonged laboratory confinement, thermal burden, task practice and circadian timing, even though the crossover design reduces many between-person confounders.

Were drug concentrations actually measured in forensic blood?

No postdose quantitative chemistry was reported for modafinil or galantamine in legally collected human venous whole blood. The study’s pre-enrollment blood and urine assessments were medical screening; these are not postdose serial concentration measurements. No centrifuged plasma or serum drug concentrations, urine metabolite kinetics, oral-fluid drug concentrations, sweat-patch results, hair segment concentrations, vitreous concentrations, milk drug concentrations, or tissue assays were reported as direct research endpoints.

The biological matrix matters if forensic testing is later introduced in a specific case. A chemically assayed venous whole-blood concentration is not interchangeable with a theoretical oral dose, a serum level, a plasma result, a medical screen, or an instrument-measured speed in a simulator. Because this study provides no original pharmacokinetic samples, it cannot validate a concentration-to-flight-control threshold, a time-since-dose calculation, or an individual impairment diagnosis. Clinical pharmacodynamics and analytical toxicology ask related but distinct questions.

What are the most important limits on using these results in court or operational policy?

Only twelve men were studied. They were young, screened and generally healthy; they were not a representative collection of patients with sleep disorders, active airline pilots, professional drivers, older adults, or people using multiple drugs. The treatment doses were specific, and repeated dosing, tolerance, medical comorbidities and chronic medication histories were not tested. The fixed-base flight scenario was comparatively simple and lacked the decision demands of adverse weather, complex instrument procedures, emergency planning and multiple simultaneous hazards.

Statistical power, multiple environment comparisons, learning effects, the timing of wakefulness, the intervening thermal exercise procedure, and missing quantitative drug exposure all limit transfer. Original numeric source material also requires caution about labeling consistency; that quality-control issue is discussed in the separate private scientific QA rather than introduced as an unsupported replacement value in the clean article. Nothing in these data tests real-world accident frequency, a named pilot’s skills, a driver’s last medication administration, a legal intoxication threshold, or specific incident causation.

How should a forensic reviewer handle a drugged- or fatigued-operator allegation?

Begin by separating questions. One question asks what chemical, if any, a validated assay identified in a properly collected specimen. Another asks which functions the individual actually performed and what impairment was observed at a particular time. A third asks whether there was a causal contribution to the alleged event after accounting for alternative explanations. This twelve-person simulator trial directly addressed only a narrow portion of the second category, using controlled group observations without real aircraft incidents or quantitative drug concentrations.

If a biological sample exists, review the specific confirmatory method, analyte, matrix, analytical range, collection time and chain of custody. For prescribed drugs, interpretation of timed observations may also require concepts familiar from therapeutic drug monitoring, but an ordinary therapeutic range is not an aircraft-fitness boundary. Assess sleep records, duty schedule, training, operational demands, coexposures and the actual event. This experiment cannot turn a pill dose into an individual causal verdict.

Frequently asked scientific questions

1. How many independent participants completed the study?

Twelve healthy young men were studied. Each person attended three treatment sessions, so the dataset contained repeated observations on twelve people rather than thirty-six independent patients or pilots. Flight segments and telemetry points are repeated observations, not additional biological participants.

2. Was the administered substance methamphetamine or dexamphetamine?

No. The treatment conditions were a single oral 100-mg dose of modafinil, a single oral 10-mg dose of galantamine, and placebo. Galantamine is an acetylcholinesterase inhibitor, not an amphetamine-type stimulant. The study cannot answer effects of methamphetamine or dextroamphetamine on flight performance.

3. Did either drug make the participants fly safely in actual aircraft?

No actual flights were performed. The outcome was maintenance of a target altitude, speed and heading in a fixed-base simulator. An actual aircraft, unexpected hazards, licensing evaluations and accident outcomes were not tested. A limited simulator result cannot certify safe real-world aviation operation.

4. Was the modafinil altitude result statistically significant?

The reported across-environment modafinil altitude comparison had p = 0.053, which did not reach the study’s conventional 0.05 threshold. The source also reported no significant altitude differences between the medication arms in each individual flight environment. Do not convert a near-threshold value into a significant medication benefit.

5. What was significant for galantamine?

Galantamine sessions showed an across-environment altitude difference, χ²(2) = 10.9, p = 0.004, and the report described a selected speed difference versus placebo, p < 0.001. Those findings concern specific statistical contrasts. They do not establish that every flight metric improved or that the medication corrected the effects of sleep loss for every participant.

6. Were modafinil or galantamine measured in human whole blood?

No postdose quantitative modafinil or galantamine concentrations in forensic venous whole blood, separated plasma, serum or other chemical matrices were provided. The reported milligram values are nominal oral doses. No individual elimination time, threshold blood concentration or chemical impairment rule can be derived from them.

7. Does a nonsignificant earlier-versus-later simulator contrast prove full recovery?

No. A null hypothesis test that does not find a difference is not a formal equivalence demonstration. Appropriate restoration or noninferiority claims require explicit margins, sufficient statistical precision and well-defined rested reference conditions. The authors did not establish a validated personal recovery threshold for aviation fitness.

Original primary research and evidence boundary

Original study: Lewkowicz R, Przewodzka A, Łaszczyńska J. Flight Performance Following Administration of an Anti-Fatigue Stimulant During 27 Hours of Continuous Wakefulness. Polish Journal of Aviation Medicine, Bioengineering and Psychology, 28(2):5–18. https://doi.org/10.13174/pjambp.23.05.2025.01. This is an independent scientific account of the specific original experimental endpoints, not an expert certification of an individual’s fitness, drug concentration, intent, crash responsibility or admissibility of evidence. The original article’s figures are not reproduced. The contextual same-site links concern general method interpretation and are not evidence that the original flight experiment measured a forensic drug specimen.